Kalzumeus Software

Archive | AdWords RSS feed for this section

How To Do A Seasonal Promotion for Your App

Posted on October 14, 2009 by Patrick in AdWords, marketing

Many software/web-app developers naively assume that demand for their product is roughly static year-round, despite the fact that this is true for very, very few industries. Retail (in-a-box) software lives and dies by its Christmas numbers. B2B software often sees spiky behavior around the accounting periods and business cycles of its target sector. If you take a look at the data, it is likely that your business also has hot and cold periods — so how do you exploit that?

For example, I sell software which makes bingo cards for elementary schoolteachers. Since a huge number of my customers use bingo as a fun diversionary activity rather than a regular instructional tool (nothing wrong with either approach, incidentally), I get big spikes of interest around holidays. Due to peculiarities of how American schools and religion relate to each other, the ideal holiday for my purposes is one where a) class is in session for b) a non-religious holiday which c) does not have an obvious history-focused lesson plan.

If you take a quick look at the school calendar for those three criteria, you might find yourself saying Halloween. And if you do, congratulations: the spike in traffic I get in October every year as a result of interest in Halloween routinely makes October my best month of the year.

So knowing that, this year I’ve done a bit of optimization for Halloween, which you might be able to adapt to your own businesses:

Mini-Sites

The keyword profile for new customers coming in as a result of seasonal interest is generally very, very different from your everyday customer. SEOing for these transient customers requires doing a bit of work, and (especially for those of us who do not have the budgets of Amazon) significantly altering the front page of the site to fit a Halloween theme (with the appropriate keywords everywhere) might be a bit too radical.

Enter the mini-site: you can take a handful of pages or, if you’re feeling generous, an entire new domain name, and then use an on-target visual design with optimized keywords to make it very, very obvious that you’re attentive to the desires of your holiday searchers. As competition for the holiday keywords is likely fairly low relative to the “head of the long tail” keywords in your niche, you can rank a mini-site fairly easily.

For example, I registered halloweenbingocards.net and put up a simple WordPress installation with 5 pages about, well, Halloween bingo. It looks much, much more Halloween-y than my main site, and empirically has been converting rather well for the first two weeks of October. (I know from experience last year, when I had no mini-site but did have pages with Halloween-related content on my main site, that the 300 or so visits a day it is getting this week are about to increase by more than an order of magnitude over the next three weeks.)

The total cost for this site (domain registration, content written by freelancers, directory listings for SEO purposes, etc) was less than $500 and probably another 10 hours ($1,000) of my time. (I wrote the first draft of it by hand, myself, before getting smart and automating/outsourcing things. You live and you learn. My next dozen ideas in this general vein are getting done much more efficiently.)

Incidentally: That WordPress design is based on a free template from WP Design, who graciously gave me permission to use it commercially (it comes with a Creative Commons non-commercial license by default, and I only realized that when writing this blog post, and promptly offered to pay for their blessing, because if software developers can’t respect license terms then we are doooooomed). It is beautiful and saved me a lot of time in designing the site, and I’ll probably be going back to them for design projects in the future.

PPC Campaigns/Landing Pages

Mini-sites require a fair bit of work and creativity to do right, but if you can’t do anything else, you can quickly create PPC campaigns for the seasonal keywords. Many businesses would use a holiday as a wonderful excuse to do a periodic sale — that is wonderful for you if it works. However, with just a little work you can respin your existing content into an on-target landing page and then toss up some AdWords at it.

AdWords optimization is generally a little outside of my comfort zone, but here’s what I did:

I ran a report on what URLs were showing my ads on the Content Network, and eyeballed them for the ones which were specific to Halloween.
I banned those URLs from showing ads in my default ad group.
I created a new ad group for Halloween-themed ads, and put the ads into it as Managed Placements, plus added in some Halloween related keywordery.
I created some Halloween-themed creatives, to catch the clicks on the Halloween themed pages that my ads would be showing on. For example:
I re-used a quick variation on my main AdWords landing page, replacing the usual content with a bit of Halloween content.

This whole process took less than 15 minutes, mostly as a result of futzing around in the AdWords interface, as the landing page was already pre-written. When I spent some time reworking my default landing page a few months ago, I linked it to the I-can’t-believe-its-not-a-CMS that drives most of the content on my website. This lets me re-use any of my pre-existing content (bingo card word lists which I pay my freelancer to write for me) as a landing page just by changing the URL for it — for example, if instead of http://www.bingocards.com/lpc/halloween I were to write http://www.bingocards.com/lpc/christmas it would suddenly be a Christmas-oriented landing page.

I strongly recommend that you have some way to quickly generate landing pages like this — if not a custom-built CMS lurking in the background, at least consider having a pre-made template which can be customized and uploaded in a few minutes. This lets you throw up a quick seasonal campaign anytime you get the inkling to — I don’t know if Dropbox does better on Boxing Day but if you can launch campaigns this quickly there is no reason why they can’t find out in the time it takes to brew coffee.

Work Your Mailing Lists

Many software companies offer interested people the opportunity to opt-in to receiving mails about special promotions & etc. (I actually didn’t do this for almost three years, because I was worried about being spammy. Then I dipped my toe in the water and found, to my surprise, that not only can you do email marketing in a non-spammy fashion, some people actually enjoy it so much that they’ll email you to ask why the September newsletter hasn’t arrived yet. Really!)

For example, when folks sign up for the free trial of Bingo Card Creator, they’re given the opportunity to opt-in to a monthly-ish newsletter. These folks are solid gold to talk to: all of them have used my product and expressed an interest in hearing about how it can make their classes run smoother. Halloween is the perfect opportunity to get in touch with them — to hum a few bars, “Hey guys, do you have a Halloween activity planned yet? Oh, you were too swamped and are putting it off to the last minute? How about bingo, using that software you already tried out this summer? It is fun, will fit in your class time, lets you use the candy you were planning on distributing as marketers, etc etc sales pitch.”

For bonus points, you can re-use or re-adapt mini-sites or PPC landing pages for your email audience. I won’t be doing 100% re-use, since folks who have signed up to the free trial before are at a different point in the sales cycle than folks clicking on those PPC campaigns: they know the software works for their needs, they just haven’t seen $30 worth of value for it yet, so they need to be told how much hassle it is going to save them in late October. Folks coming in from the PPC campaign, by comparison, know they want to play bingo in late October, but they don’t know that my software can help them do that and that I am not some evil scammer on the Internet. (Fear and distrust of the Internet and the evil virus-spreading Spammy McPhishertons on it runs rampant in my market, so I spend significant efforts trying to demonstrate that I’m on the up-and-up. That would make a good topic for a blog post, actually…)

You can also offer incentives to responding to an email. Discount coupons are traditional, but as web-savvy software engineers we can be much, much more devious. If you have a game or StackOverflow-type site you can offer a time-limited themed achievement. I’m planning on offering to give them a Thanksgiving activity for free if they log in before October 31st. (If that sounds wacky: giving them a Halloween activity for free would destroy any need for them to pay me money right now, but giving them a Thanksgiving activity for free creates a powerful incentive to come to your site and listen to your best possible Halloween sales pitch. If they weren’t going to log in in October anyhow, giving them a freebie for Thanksgiving costs me nothing, since they were vanishingly unlikely to decide “Hey, I think I’ll log into that dormant account and whip something up for Thanksgiving, fall in love, and decide to purchase it.”)

I want to expand on that last idea a little bit, because it is so powerful for software developers: any IP you create can be replicated essentially indefinitely for no marginal cost, much like your main product. However, customers don’t see the world that way — they live in a world where scarcity is a reality, and accordingly perceive value in things they don’t have but might want. Thus, you can trade your customers access to IP in return for going further towards a conversion with you, and scale this offer across all your customers. The economics of this are staggeringly efficient compared to e.g. PPC advertising, and you’re giving the offer to folks who are already pre-disposed to liking you, so the conversion rates should be much higher than similar techniques aimed at “cold” prospects.

Both making the offer and fulfilling it can be automated, so it costs a static amount of labor no matter how many customers you have and how many take you up on the offer. If you’re smart and make this process repeatable, it actually takes less time every time you do a similar promotion, since you’ll have the infrastructure already and all you have to do is create or buy the premium.

Trick or Treat

Well, there you have it — three simple techniques you can adapt to almost any holiday or seasonal promotion. I hope they got the juices flowing a bit. If you’ve got any fun ideas in the same vein feel free to leave a comment — I love brainstorming with folks.

Look for a post in early November on how these techniques actually ended up working for me. Early results look rather positive.

Landing Page Design Tips

Posted on August 6, 2009 by Patrick in advertising, AdWords, web design

Recently, I finally got around to making a proper landing page for Bingo Card Creator. It has been fairly successful for me so far, increasing conversions markedly (my historical average for AdWords is about 18%, that page converts at about 28%) and decreasing my cost per conversion.

Here’s what I think I did right:

Focus On A Single Goal

The stock Bingo Card Creator homepage, which has been the landing page for all of my advertising since time immemorial, has to serve a lot of purposes at once. It signs people up for the trial. It drives people to the download. It sells the benefits of Bingo Card Creator. It provides an entry-way into support. It lets people log in. It funnels around link juice for SEO purposes. It, like the prototypical uISV, wears fourty-seven hats.

Landing pages should have one goal. My new one aims to capture a sign-up to the trial of the online version of the software. That is it.

To accomplish this:

Strip out all navigation which conflicts with The One True Goal.
Eliminate visual distractions from The One True Goal.
Make it absolutely, blazingly visually obvious what The One True Goal is.

If you guessed that the goal for the page is clicking either the giant purple button or the single visible hyperlink, you win!

Include Trust Markers

Bob Walsh has gone over this a million times referring to his MicroISV Sites that Sell book: testimonials work. I have never really used them to their full potential, but I have diligently collected a little notebook (with permission, naturally) from my 1,800-odd customers. This gives me a bit of a palette to work with. While normally I personally prefer hitting customers with a few testimonials at a time, like this wonderful example, I didn’t have too much space to work with. So I picked a single testimonial, lightly edited for length and content, and bolded the important bits.

I cannot stress enough how much people do not read on the Internet. Accordingly, if you want to maximize your sales, you should adapt to users that:

skim more than they read
feel more than they think
enjoy bad self-referential humor (OK, this one is optional).

Seriously speaking: bold the bits of the testimonials which sell your software. Cut the bits that don’t, as long as you can do so without making the testimonial say something the customer did not. They are not professional marketers and should not be expected to write like them. You are, so give them a minor hand. (Polishing it until it is too good is another danger. Testimonials should remain authentic.)

It should go without saying but in addition to reflecting on you, a testimonial reflects on your customer. Even if you are making them largely anonymous, extend them the courtesy of copy edits to make their words suitable for public distribution. If I never see [sic] in a testimonial again it will be too soon.

Don’t Forget the Guarantee

I don’t feel like elaborating on this, other than to say I once walked out the door without pants on. That is a pretty serious oversight, but it still isn’t as bad as forgetting your trusty money-back guarantee.

Address Your Shadow Audience

Google likes to see certain things on landing pages and hates to see others. You really don’t want to tick them off. Most of my readers are honest businessmen, which automatically saves you a lot of the headaches, but there are a few subtleties:

Trust markers like a privacy policy, return policy, etc go a long way.
Don’t look like an affiliate. Google is very ambivalent about affiliates — they make billions off of them, but to the maximum extent possible Google would rather disintermediate the affiliate, because they feel that a) it provides a better customer experience and b) Google should be getting the money the affiliate does.
Landing pages have a tendency to repeat large swathes of text, which could cause duplicate content problems for your site. This is particularly acute if you build them on an industrial scale (about which I will have more to say later). Save yourself some pain and either exclude your landing pages from the index using robots.txt or include the following code in your header somewhere:

<meta name="googlebot" content="noindex, noarchive, nosnippet">

Technically speaking there are more search engines than Google to worry about. Hah. Sorry, where was I.

Provide Continuity

Marketers sometimes say “information scent”, but I like continuity: seeing your ad, clicking on your ad, reading your landing page, and then interaction with The One True Goal should be one continuous flow, free of jarring transitions and mental static. For example, you should have your ad call to action anticipate the call to action text on your landing page. Thus:

Notice how I’m focused on the proposition of “quickly”, with the call to action “Try now, no download required!”

That matches my call to action on the landing page:

This means that for a large portion of my visitors, I get the click to the purple “for free” : they’re already mentally invested in trying now, no download required. That is why they came here. So I don’t have to convince them so much with the other text on the page.

Of course, that doesn’t help much if I lose them at the signup form. Candidly, my signup form could use some work — it performs much, much better than my download/install rigamarole ever did, but it severely needs a graphical upgrade and some interaction design TLC. However, there is one trick to it that I’m fairly happy about. I’ll share that with you guys this weekend, as it is a little subtle and I want to collect a few more days of data to show you that it actually works.

Preview of coming attractions: I have often said that paying a freelancer to write content for my website was probably my best ROI ever. This resulted in over 700 pages which are very responsive to exactly what searchers want. What if it had also resulted in 700 ads?

My AdWords Are Turned Off [Edit: Resolved]

Posted on August 1, 2009 by Patrick in AdWords, customer service

[Edited to add: My campaign returned to mostly normal on August 4th or thereabouts. I still don’t know what the heck happened, but I’m happy things are normal. I sent Google a second support request, but that may or may not have had anything to do with the fixing: it was not responded to. I also added two additional ads (which both triggered a human review), upped my bids, and in general tried to flag the system to say “Hey system, new stuff here!”, so that may or may not have had something to do with it. My current feelings about Google, on a 1 to 10 scale: performance of platform: 8, superiority over competing ad options: 10, support responsiveness: 3. That 3 is generous, based on my completely unevidenced thought that it is possible someone inside Google actually saw the problem and intervened. At the moment, though, I have as convincing evidence for the existence of Google technical support as I have evidence for the existence of Santa Claus — luckily for Google, in my heart of hearts, I still believe in Santa Claus.]

High on the list of things to do today was try my hand at making my first landing page. I have recently started selling my product as an online web application in addition to downloadable software, and wanted to see if focusing pages on one or the other would improve my conversion rates.

Naturally, I had to log into AdWords to create the alternate ads to test this landing page against my usual ads, which just dump people at my homepage. Since AdWords is very fire and forget for me (yay, Conversion Optimizer) I don’t log in more than once a month. Thus, I hadn’t known that since July 23rd they’ve been ratcheted down from “most schoolteachers are out for summer so you’re only getting 10k impressions a day” to “almost turned off so you’re getting 20 impressions per day”. Since AdWords account for a large portion of my sales this is, to be blunt, absolutely terrifying.

Delving Deep Into Black Magic

AdWords includes a variety of automated means to diagnose why your ads aren’t showing. They provided the very useful advice “It is because you haven’t input a valid means of payment yet.” Google, the charge for $200 yesterday went through as swimmingly as the last $12,000 have gone. You have my freaking credit card on file.

This left me with another alternative, contacting AdWords support. Google really buries the email us button (behind useful automated diagnostics like the above, and equally helpful FAQ entries) because Google hates dealing with customers. We’re little annoying things that don’t scale well when we can’t be handled perfectly algorithmically like all the world’s information (TM). Stories about how their support is incompetent and outright hostile to speaking to customers are legion.

I say this as someone who is a fan of Google. Full disclosure: I am a case study about AdWords’ effectiveness. Literally. When the system is operational it rocks my socks off, but when it goes off the rails you are screwed.

They’ve improved response times since the last time I used them. It took me less than an hour to get a response. (Aside: Last time it took almost two days to hear from them when I wrote in to complain about one of my ads being disapproved. That time I got copy/pasted a portion from their TOS implying that St. Patrick’s Day bingo cards were gambling paraphenelia. It took three back-and-forth emails before I convinced them that I am, in fact, not in the gambling business and if they read that cough Google case study cough they would see that. But the ad missed St. Patrick’s Day, which rather sharply limited its commercial utility to me.)

AdWords Support: Like Talking To A Markov Chain

My email (which was limited to 512 characters, because Google apparently after indexing the entire freaking Internet Google didn’t have enough hard disk space in BigTable to save bug reports of a useful length) explained who I was, what behavior I had seen (ad impressions down by a factor of a thousand), what behavior I had expected (ad impressions not down by a factor of a thousand), what factors I had already ruled out (disapproved ads, sudden CTR decrease, sudden conversion rate decrease, payments issues, quality score issues, etc, etc), and what resolution I wanted (a fix or cause that I could fix myself).

Here’s what the Googleplex Markov chain spit back:

It looks like you have some questions about your ad rank.

Translation: I did not even pretend to read your email, or else am unaware that the Content Network has no concept of ad rank per se.

In the future, you may find that the quickest way to find the answer to your question is
 through the Help Center at https://adwords.google.com/support/?hl=en_US?utm_id=hc. Or, try the
 AdWords Help Forum at http://www.google.com/support/forum/p/AdWords?hl=en?utm_id=aut, where you
 can share information and exchange ideas with other advertisers.

Translation: We really hate when you email us. Please don’t.

Please note that it's possible your account may be under review. As you
may be aware, we periodically perform these reviews to ensure the highest
quality ads, verify billing information, and maintain general account
security. Per our Terms and Conditions
(https://adwords.google.com/select/tsandcsfinder), ads can undergo review
at any time. If your account is under review, we will get back to you
shortly.

Translation: You may be evil. We don’t do evil, hence, we don’t do you. You may not be evil — that would be a good thing, but we still hate talking to you. If you’re evil, we kind of have to talk to you. If not, whee, we don’t have to talk to you, so don’t be evil.

For your convenience, we've listed some relevant information below:

Translation: Here are twenty random links from our knowledge base. Despite our expertise at ordering the world’s information, they are neither relevant nor even plausibly related to your inquiry, and most are an insult to your intelligence. They won’t resolve your problem, but possibly they’ll stop you from trying to email Google again, which is our ultimate goal.

The uselessness of some of their suggestions beggars belief. “Why do the same ads show on different pages?” Well, I didn’t click, but I’m going to guess that the reason the same ads show on different pages is because I paid $10,000 to have Google put my ads on their Content Network. If my ads were not showing on different pages, that would be a problem. Oh, guess what — they aren’t!

Sincerely,
The Google AdWords Team

Translation: You weren’t even worth enough of our time to have one of our $6 an hour Indian callcenter employees sign her name to this.

What I Think Went Wrong

While waiting for the (as expected) useless reply from Google, I played forensic investigator with my stats from this year and last year. Everything looked nominal — in particular, Conversion Optimizer is sensitive to conversion rates so if your website suddenly stops converting, you’ll find your ads turned off in short order. (That is what you’re paying them for, after all.) I verified that my trial download was still converting as expected. Yep, code present in pages, yep, code evaluating, yep, conversions reported in AdWords so Google must be getting the data.

Then I tried it for the online version. The online version scores a conversion any time someone signs up for the trial and logs in successfully (that should be automatic on signup, but you never know, particularly with bots). To prevent the same person from getting scored as a conversion every time they log in (which would cost me gratuitous amounts of money), I have Rails set up so that only on their first login after signup they get sent to a welcome action. The welcome action, in addition to doing a bit of housekeeping, is supposed to set the @welcome instance variable so that the post-sign-in page knows to display the conversion tracking Javascript.

def welcome
#Snipped a portion not relevant to this discussion.
@welcome
dashboard
render :action =&gt; 'dashboard'
end

This is a Rails idiom that runs the dashboard method and then renders the dashboard action, so that the welcome-specific code gets executed and then the shared template gets called. Did you spot the bug? Yep, that @welcome really needs to be initialized to true. I’m not quite sure how this got by testing but it would be easy to miss — the page will display fine and all the functionality works perfectly, it is just that Google fails to get a wee little ping to credit my account with a conversion. Which I didn’t notice since my download page continued to ping as usual.

I’m not even sure this was the problem. Even with the online trial siphoning off conversions that should have been credited, my conversion rate for July was close to historical norms (the online trial was largely adding conversions, not replacing them). The issue with conversion siphoning in Conversion Optimizer is if you don’t score conversions then your apparent Cost Per Action goes up, and if that routinely exceeds your maximum desired CPA then they won’t bid for you anymore. However, my CPA stayed at the usual level for the 3 weeks where a portion of conversions weren’t scored, prior to my ads not showing.

What Next

Well, since I fixed the bug, I’m sort of in a wait-and-see mode. If their algorithms auto-correct me back to my previous status, then life is peaches and cream again. I also went ahead and scheduled a few new ads with my new landing page, which should start with very little “history”, positive or negative, so they’ll largely not be effected by this. I know Google heavily biases their historical computations to consider recent history rather than overall account history, so I’m a little worried about making a full recovery.

And I’m blogging about this, obviously, because I want anybody else in the same situation to have some support which is not a Markov Chain.

Google Features Bingo Card Creator

Posted on January 24, 2008 by Patrick in AdWords, marketing

I’m sorry, this post was due a week ago but I have been a combination of sick and busy.

Regular readers of this blog might remember that, around September, I started using Conversion Optimizer on my Content Network campaign and had a lot of successwith it. My success came to the attention of the project manager for marketing Conversion Optimizer, and after a bit of discussion we agreed to cooperate in the production of a case study. Google published that case study fairly recently — it doesn’t include much that you didn’t already read in the above posts and the followup, but it might be interesting to see how their take on it is subtly different than my take on it. You’ll note, for one thing, that they don’t mention that the prevailing view of the the Content Network is that it is a “hive of scum and villainy”. I can’t imagine why.

I don’t have definitive knowledge as to why Google chose to select me for the case study. I have theories, though. The first was that I took an early lead in staking out the topic area when it was released, and it made excellent sense for Google to talk to someone who was already ranking high for their own product name. It couldn’t have hurt that I was largely positive and transparently willing to share the exact statistics, which is great for someone who doesn’t want to waste their time going through 15 levels of corporate officers and ultimately failing to get the go-ahead to release the numbers.

I also think that authenticity mattered. On an Internet which is chock-full of scam artists and folks wanting to make a fast buck at others’ expense, you can stand out pretty easily just by bucking the trend. If you have 15 competitors and you’re the only one who is radically transparent about your business, then for many opportunities you have no competitors. You’re sailing in a Blue Ocean, and you’re the only ship there is, from one side of the horizon to the other.

I think uISVs should make use of this, not necessarily by being radically transparent, but by running businesses which stick out because they are meant to stick out. The software can be cloned, the keyword lists duplicated, the advertising imitated, but no one has come up with a copy machine which can Xerox the soul of an organization that has one. If your differentiating factor is a sense of humor, fanatical devotion to customer support, personal expertise in your domain, an emotional connection to your niche, or just the bare fact of being the little guy, that helps you stand out in a sea of mediocrity.

OK, that is enough of the theorizing. So, further thoughts on Conversion Optimizer: it still rocks. I pay about 24 cents for trial downloads from the Content Network these days and, after someone corrected my mistaken impression that you couldn’t use it for search, my search campaign has had a solid three months below 30 cents (which was my gold-standard for execution back when I was hand optimizing). This has also freed up well in excess of four hours a month, which I have been using to convalesce (boo!) and plot my next improvements to my website (yay!). I really can’t recommend it enough.

Full disclosure: Google gave me a laptop bag as a thank-you gift for participating in this case study. Some folks might think that compromises my objectivity. I don’t think it compromises it nearly as much as helping me make a few thousand bucks, because that buys me an awful lot of laptop bags. (Although those wouldn’t get me stopped at customs — “Oh, you work for Google!? ARE YOU FEELING LUCKY?! Sorry, sorry, I tried to make a joke. Was it funny?” Frankly, we need more good bad jokes to pass the time at airport security.)

Exploiting New Niches

Posted on November 24, 2007 by Patrick in AdWords, blogging, SEO

I’m currently the #1 result on Google for the [conversion optimizer] search which is not actually controlled by Google, as a result of a pair of posts on the subject.

That is largely the result of a combination of a niche which was new, my early adoption of it, and radical transparency. There are, approximately, a billion Internet marketing and search engine watching blogs out there. Many of them are much, much higher profile than this blog. Many of these covered the launch of Conversion Optimizer, but I was (to my knowledge) the only person who backed up the first impressions with numbers, and as a result I attracted just a few more links than the next guy and, wham, the search engines think I’m the expert on the topic.

I mention this for two reasons: One, it is quite useful to know how to convince the search engines you are the expert on a particular topic. For niches which are brand spanking new (say, hmm, Blackberry spam filters after the introduction of the Blackberry), the combination of there being zero pre-existing links, less than a full Internet of competition, and massive first-mover’s advantage means that you can snag the top spots quite easily if you move reasonably quickly with something compelling. It’s not just enough to be there first, you have to be there “firstest with the mostest” — I think my first post on Conversion Optimizer was weeks after launch but the “real numbers” hook is extraordinarily more compelling than the “news you already read from Google’s blog” hook.

After you’re already on top, you’ll probably stay on top, because the guy on top becomes the canonical result to refer to the subject when anyone else just needs to introduce it. (This phenomenon is described in Filthy Linking Rich, probably the most worthwhile article from 2004 for a business owner in 2008.)

Aside from it being a useful business skill to learn how to position yourself as the expert on an emerging topic (and, for what it’s worth, I’m hardly the expert on this subject, just expert in the eyes of the computer algorithm that people trust to identify experts these days), this opened up a nice opportunity for my business. I hate to be coy, but it will be another week or so before I can say exactly what it was. For now, I just wanted to get this post timestamped so that I can refer to it for a before-after comparison in the wake of the announcement I expect to be making.

Conversion Optimizer: AdWords, Done Right

Posted on November 10, 2007 by Patrick in advertising, AdWords, bingo

I have been using Google Conversion Optimizer since late September. The program has been extraordinarily successful for me, enabling me to double revenues (and profit margin) in a month. This post recounts my experiences.

Who I Am: I sell software which creates bingo cards to teachers. The goal of my advertising campaign is to get them to download a free trial. Some portion of those who download a free trial purchase the full product, making me money.

What Content Network Is: Conversion Optimizer (hereafter CO) is a way of bidding on the AdWords Content Network, which is all those little ads you see plastered on non-Google websites. Typically, on the Content Network you write up a list of keywords, write up a maximum bid per click on each keyword, write some ads, and then get charged as the ads get clicked on on any site Google thinks matches your keywords.

How Converison Optimizer is different: With CO, you don’t specify how much a keyword is worth to you. You specify how much a conversion is worth to you. Google then guesses, based on your previous history, what CPC (cost per click) price will achieve your CPA (cost per action) goal at any particular website, and bids accordingly on your behalf. I like to think of this as pseudo-CPA bidding, since its really just a wrapper around CPC bidding. Google has a beta CPA product that I haven’t tried yet, and probably never will, because CPA bidding has some severe issues for publishers which will make it perpetually less popular than CPC bidding. There aren’t enough publishers in my market interested in CPA bidding to make it worth my while to start a campaign.

My Experience With The Content Network Prior to Conversion Optimizer:

Endless frustration pretty much sums it up. When I started advertising on Google last September the content network was on by default. At the time it was filled with Made for AdSense (MFA) sites, which scrapped content and existed only to get clicks. They got clicks, but the traffic I got was extremely low quality, and they drained my AdWords budget dry. I was advised by knowledgeable folks at the Business of Software boardsthat the Content Network was a hive of scum and villany. I turned off the Content Network, advised everyone else to turn off the Content Network, and went on building my business for the better part of a year.

Why I Came Back to the Content Network: A professional peer who produces language learning software told me he had had excellent success with the Content Network. I am skeptical about most people opining on the Internet but I have a good deal of trust for him and, moreover, I trust data. Google had in the interim cleaned up some of their click fraud problems and shared Placement Reports with advertisers, which let you see which sites were running your ads and what their individual conversion rates, CPC prices, click through rates (CTR), and whatnot were. With that data, I thought I could make something of the Content Network.

So things went well? No, actually, they sucked for quite a while. I got trapped in an endless cycle of checking my Placement Report every day, banning whatever flavor-of-the-moment site was abusing their own visitors to misdirect folks into clicking on the ad by disguising it as content (n.b. the site owner featured there says it was an accident, and I believe him, but I was getting a new site using a similar technique two to three times a day), and losing the money I had allocated. I have a day job at a Japanese company, with a 3 hour commute. When I get home at 11 PM at night, I don’t want to sit down and have to immediate check AdWords to see if my pocket is getting picked yet. I was right about to turn off the Content Network again when CO debuted. I jumped at the chance, and became one of their earlier advertisers.

Anatomy of a bad site to advertise on:

Let me pull you an anonymized entry out of my Placement Report, and compare it to my averages. These images are digitally edited composites of my AdWords status screens — I haven’t changed any numbers, but I have removed some to make the narrative clear. If you are reading this on WordPress.com and not on a feed reader, the image is truncated — click it to see the full size.

As you can see, the bad site has a CTR much, much higher than my average CTR. This is because their site is designed to funnel their own visitors onto the ad, which is disguised as content. Notice the conversion ratio is terrible. This is exactly what we would expect — since they were tricked to click on the ad, they have no real interest in downloading a free trial of my software, and click the back button to try to find what they were looking for on the dishonest site. This drives my Cost Per Conversion through the roof (incidentally, my goal is 30 cents, and break even is about 60 cents).

Multiply that one dishonest site by about 20 or so, with more showing up every day, and you’ll understand my frustration with the Content Network.

Anatomy of a Good Site on the Content Network:

Here, on the other hand, we have a good site on the Content Network. They wrote a page with unique, original content which is useful to folks in my niche and which got them in the right mood to need Bingo Card Creator. Then, in a perfectly honest fashion, they had ads after their content that folks could click on if they were interested. You’ll note that while 3.62% is a very good CTR it isn’t outlandishly, unbelievably good like 20%. You’ll also note that their conversion rate is high (18%, a bit above my average for content network, which underperforms organic and paid search), and that I am actually profiting from this site. Substantially.

Heck, I’ll out the good site: it is About.com’s page on Halloween bingo. Thank you, About.com, you made me about $250 in October. I salute you.

Where Conversion Optimizer Comes In:

You’ll note that the major observable difference in the numbers between Bad Sites and Good Sites is that Cost Per Conversion value. Heck, from my point of view, that is the major reason why they are bad: if I pay $3 for a trial download, which converts at approximately 2.5% to a $25 purchase, I am spending $120 to make $25. Not a good idea, and I’m supporting the pollution of the Internet in the bargain. If I pay 30 cents, I am spending $12.50 to make $25, and supporting good content creation. Good idea!

Conversion Optimizer makes me bid on Good sites, not on Bad sites. And that is why it works phenomenally, phenomenally well. If I advertise on Bad site, their CPA price is through the roof for a day, and Conversion Optimizer silently and automatically makes sure I never send them an impression ever again. Conversely, if I bid on a Good site, Conversion Optimizer analyzes how profitable that site is and steadily ups my bid so I can capture as much traffic as possible while remaining profitable. This is why all of my Good sites with any volume hover right around the $.30 cents CPA. (My bid is 35 cents.)

How Conversion Optimizer Is My Own AdWords Manager:

My market is very seasonal, and ideally I would be writing a new ad, ad campaign, and landing page for every major holiday, the start and stop of the school term, and whatnot. But who has the time? I was working 60 hour weeks in late October and, while I made some effort to capture the coming Halloween traffic, I couldn’t spend many hours on optimizing the heck out of my pages.

But why write landing pages when the rest of the Internet has already done it? There are many, many pages out there extolling the fun of Halloween bingo and lamenting how hard it is to actually make the cards for it. About half of the ones in the Google top 10 for relevant queries run AdSense. They saw major traffic as Halloween approached, and that traffic was highly, highly motivated to play Halloween bingo… and my ad was in front of all of them, NOT because I had had the brainstorm “Hey, I should get the Advertising Department working on a coordinated campaign to shove 80% of my advertising budget into Halloween bingo” but because a soulless algorithm decided “Hmm, Bingo Card Creator is having its best return on investment on these sites. I’ll put most of his advertising there.” Its like my own private campaign manager… who works for no salary. Lets have a scary Halloweenish Mwahaha! for that.

The Before and After Picture:
CTR 2.23% vs .5-1% (up by a factor of four. This makes it easier for me to get cheaper clicks, because AdSense prefers to show ads which generate income rather than just having high bids.)

CR 18% vs 20-25% (not a significant change)
CPC 6 cents vs 9 cents (costs down by about a third, nice enough, but wait for it…)

Profit $500 vs $10 (Note that since my business previously profited about $700 a month, that is sort of a nice boost)

Wait a second, how did it make that much?

By scaling. Conversion Optimizer, unlike my hand-edits to AdWords, just kept scaling as I tossed more money at it. I ended up with an AdWords bill in the hundreds, to be sure, but I ended up with a 100% ROI on the hundreds, instead of ending up with a 20% ROI on my previous tens.

Is it continuing?

The Halloween spike has subsided, and the Optimizer isn’t spending all of my budget allocation every day anymore (sadly), but Thanksgiving is right around the corner… And I don’t have to do anything to prepare for it this year — the clicks will come straight to me, from whatever the Optimizer decides are the most worthwhile Thanksgiving bingo sites on the Internet as measured by their actual success in making me money.

Bonus Points:

It is difficult to tell if this number is totally accurate, because Google’s conversion tracking is sometimes lackluster, but I believe my trial-to-purchaser conversion rate on Conversion Optimizer-generated trials was about 4% as compared to my typical 2-2.5% number. I attribute this partially to Halloween seasonality and partially to Conversion Optimizer finding me only the most motivated prospects, instead of folks who in aggregate didn’t really need Bingo Card Creator. Obviously, increasing that conversion rate by 60-100% has a corresponding direct impact on my ROI.

Any bad news?

You can’t use Conversion Optimizer if your Content Network campaign gets less than 300 conversions a month. When I spoke with one of the product managers he told me that was unlikely to change, as there are statistical significance issues below that number and the engineering team doesn’t think they can push it any lower. I feel for you guys with CPAs in the $10 range, as I wouldn’t want to have to risk $3,000 on an unproven and relatively new product either.

There is also some tin-foil-hattery about giving Google enough information to guess where your profitability is. In the end, I think as long as I get trial downloads at 30 cents in quantity I could care less if I gave up all of my customer surplus (econ term) from a smaller number of trial downloads achieved at under 30 cents each. As you can see, the absolute number of dollars the campaign puts in my pocket exploded with Conversion Optimizer.

In a Nutshell:

Conversion Optimizer just made me a whole lot of money. I recommend you try it out if it sounds like a good thing for your business’ unique circumstances.

Like this post? This blog contains nothing but real experiences with real numbers about advertising, SEO, customer service, and every other aspect of running an online business. If that sounds interesting to you, sign up for the feed. If you know someone in the market for software to make bingo cards, I’d greatly appreciate a mention.

New Adwords Feature

Posted on September 25, 2007 by Patrick in AdWords

I’ve been patiently waiting for a while for Google to finally come out with CPA (Cost-Per-Aquisition — essentially, affiliate sales) AdWords for a while now. They’ve got it in beta testing but that doesn’t help me. The idea would be pay folks $10 or so for every copy of Bingo Card Creator their site *sells* as opposed to 10 cents every time someone clicks on over to me. The idea is that that this would reduce the incentive to use scummy tricks to incite clicks.

Anyhow, Google recently released a kinda-sorta CPA feature. They call it the Conversion Optimizer. In a nutshell, instead of setting a maximum Cost-Per-Click (CPC) bid, you set a CPA bid, and Google uses some black magic to decide what a click at a certain day, time, and website is likely to be worth to you, based on your campaign history. (You’ve got to have significant campaign history — 300 conversions in the last 30 days, which would imply you are also spending a chunk of change on Google. Given that I just got charged $200 on my card for AdWords ads and that my last charge was on, hmm, September 15th, I qualified pretty easily.)

So, being the absolute sucker for new technology that I am, I signed up my Content Network campaign for it. Over the last month my Content Network Campaign has had an average CPA of 41 cents. I’d like to get that back down to 30 cents, but for the moment for testing purposes I set my CPA bid to 35 cents.

I’ll keep you posted with how this is working out for me.

P.S. While $280 (and counting!) of advertising expenditure this month is giving me a bit of heartburn, on the plus side, I’ve just crested $1,000 in sales for the first time ever.

Deceptive AdSense Ads Worse Than Click Fraud

Posted on September 20, 2007 by Patrick in advertising, AdWords, web design

Much has been written about the dangers posed by click fraud on the Google advertising products, and how Google has taken steps to address the problem. Click fraud, however, is only one of the ways for webmasters to defraud advertisers of money. I will detail another way in this post. The technique is already widely known among webmasters who use AdSense (and, indeed, sometimes I wonder if Google doesn’t encourage it). If you’re spending money on the Content Network, you also need to understand it so that you can cut your losses when appropriate.

A bit of back story: recently, I bumped by spending on the Content Network up by 30%, to the “several hundred dollars a month” range. As you might imagine, at 9 cents a click (bingo cards aren’t the world’s most competitive niche) this means I was getting a virtual torrent of traffic. During my daily check of the summary statistics (a habit I suggest you get into after major changes to AdWords — in normal operation once a week is fine), I noticed that my click-through rate (CTR) on the content campaign had skyrocketed from 1% to 15%.

That couldn’t possibly be natural. Remember, an AdSense ad is, by definition, being shown to someone who is at least partially interested in something related to your product but has not expressed any interest in being sold to yet. (Folks on Google frequently have expressed an interest with their search queries, such as buy bingo card creator, which is why CTRs are orders of magnitude higher there.) As such, you should expect CTRs to be much lower than on AdWords ads — dropping from 8-10% for a really good AdWords ad to about .5-1% or so for AdSense on most sites.

However, site design can have a major influence on how effective a site’s ads are at getting clicked. Google recognizes this and teaches some of the tricks to optimize the ads (which, after all, makes them money): blend the ads into your site, place the ads where they are likely to be clicked, etc. However, they have an anti-fraud policy for sites which toe the line, because using certain techniques to get the ads clicked on results in non-interested surfers clicking them, and that costs advertisers money and drives them away from the service.

Since web pages are made to be scanned, anything that causes your eyes to be drawn toward an ad but away from its content causes your click-through rate to soar. One previously common tactic, which is now banned, was to line up images with the advertisements in order to suggest to visitors that the links provided explanations for the images. This resulted in quadrupling CTRs for the ads. Since the AdSense equation is

Revenue = (traffic) * (# of ad units) * (CTR) * (cost per click) * (percentage Google gives you)

that quadrupled revenues for participating webmasters. I’m strongly tempted to say “unscrupulous webmasters”, because once the visitor realizes they’ve been had they’ll be on the back button without a second thought, costing the advertiser money without giving them any chance to pitch their products to an interested customer. That is, of course, the entire point of the excercise.

So that is the old scam. Here is the new hotness: using CSS and HTML, organize your website in the fairly typical sections-broken-by-heads style. Then, optimize your CSS such that the section travels off the page, with the clipping at a common resolution (800×600 or 1024×768) happening in such a way as to cut off the legitimate content and thereby give your visitors the impression that the ad is the content promised in the section headings.

There are at least eight sites which are using this technique in the quite non-competitive bingo card niche. I have taken a screenshot of one site which I thought was iconic. (Editor’s note: After first posting this, the author of that site got in touch with me and said the placement was accidental. I have no particular reason to disbelieve him, as inspection of his other pages shows a variety of ad placements. I’m afraid that accident doesn’t explain the other sites, though. I am keeping the pictures up to demonstrate the general tactic, but have edited the remainder of this post to be less accusatory of his site in particular.) You really have to see it in full-screen glory to appreciate the effect. That screenshot is about 255kb and shows the site in default IE7, but if you wanted to be really devious you can use CSS hacks to make it work equally well in all browsers at once, using pixel perfect layouts and a bit of elbow grease. I have obscured the “branding” of the site, and have obscured the ads of my competitors to avoid associating them with it. (If you happen to be a competitor of mine, drop me an email and I will happily give you my list of sites which are using these strategies, or you can make your own as described below.)

Here is a close up on the main content area of the page. Again, you really should look at in in context — the actual CONTENT here is invisible until you scroll. Unsophisticated visitors miss the distinction between the blended links and the advertisements (which happen to have quite similar titles) and click on the ads instead of the file links. Click to see the expanded version.

Remember, the site does not actually show that content in the middle unless you scroll down to see it — and even with the content there, it is easy for an unsophisticated Internet user to click on the ads thinking they are getting the promised downloads.

And click they do. From my statistics, roughly 16% of the visitors of that page clicked on my one, single advertisement. Given there were five advertisements, a click in my niche costs about a dime, and Google splits somewhere in the general neighborhood of 50-50 with webmasters, we can guestimate their revenue per thousand visitors using the above formula:

Revenue = 1000 * 5 * .16 * .1 * .5 = $40 CPM. (Edit: The site owner suggests that he is earning $7.50 CPM for the site as a whole. I don’t have access to his console, but I think my estimate is closer than his for pages which employ this technique.)

Sorting the list of the hundreds of advertisers I am paying, and ignoring ones for whom small numbers distort results, it seems like a more typical CPM for an honest advertiser in my niche is about $2.50. So its fairly obvious why breaking the rules is so attractive — a single page with less than 1k impressions a day could generate something like $12,000 a year.

And when I say generating, I mean “taking it from the advertisers”.

Most business owners understand the economics of advertising a product, but a brief review for the peanut gallery: I sell a $25 product, of which $24 is profit. (It helps to be in software, the gross margins are quite healthy.) The primary goal of having a user visit my page is to get them into the free trial of the software, which convinces about 2.5% of them to convert (i.e. buy), getting me my $24. Thus, it is rational for me to spend anything less than $24 * .025 = 60 cents (at the margin) to achieve one trial being downloaded.

I have reason to suspect, given a year of data, that the attractiveness of my website and sales proposition should convince about 22% of interested visitors to take the trial for a spin. Given that clicks in my niche cost about 9 to 11 cents each, this gives me an average cost of about 36-43 cents per trial download (it bounces around on a daily basis). As 43 is less than 60, that means I am mildly profitable, with not too much room for error (if my conversion rate decreases to 2% and my cost per trial rises a few pennies I’m not making money anymore).

Bamboozling visitors to click on my ads hurts me more than errors ever could.

When an unsophisticated Internet user clicks on the “Create Bingo Cards” link thinking “This is step #1 of the 3 step process this website is pitching to me”, and then they are suddenly whisked to my very visually distinct site, they figure “Uh oh, something went wrong”. And they immediately click the Back Button, to try to fix the mistake. (Many of them probably click on a different ad instead, a mistake which is frustrating for them and great news for both the publisher and Google.) As a result, it wasn’t 22% of folks coming in from these ads who actually completed a trial download, no, it was about 2%. Which means that I was paying approximately $50 to get a sale of a $25 product — I guess I can make the loss up on volume?

Oh, but it gets worse: Google is very, very smart about where they show your ads. This is why they have a Content Score for the search network which prioritizes high CTR ads over low CTR ads: this maximizes money. Google’s incentive is to maximize the number of clicks while minimizing the number of impressions, because if they capture 100% of my budget then they want me out of the rotation ASAP so they can sell the inventory to another ~~sucker~~ advertiser. This unholy, and I hope unintended, alliance of Google and the publishers using this trick sucked my budget dry within the first two hours of every day. Google’s automated algorithms helpfully suggested I increase my spending by a factor of ten to compensate, so that instead of spending $15 a day to make $7.50 I would be spending $150 a day to make $75, for a monthly loss in the $2,000 range.

That Certainly Sucks. What Can I Do About It?

1) First, if you’re not in the position to routinely monitor your AdWords performance, opt out of the content network and don’t come back. The scum sites are always one step ahead of Google, by definition, and if you’re not one step ahead of them that $2,000 a month loss could be yours.

2) If you are in the position to routinely monitor your AdWords performance, use the Reporting feature in your AdWords console. The report you want is Site Placement, for the previous 7 days. Make sure you include the CTR and Cost Per Conversions columns. Then, every day, grab your report in CSV format, and run a simple script on it to report all of the URLs where the CTR is higher than a threshold (I use 4%), the number of clicks is substantial (otherwise you’ll ban a lot of mom-and-pop sites for no good reason because 100% of their 1 visitors this month clicked your ad), and your Cost Per Conversion is greater than your profit. (Almost guaranteed if you set your threshold right, because the only way to beat that threshold is to be exploiting your visitors, and exploited folks don’t make happy customers.) Then, take any domain which appears on this screen, and add it to your banned list.

I am a Cygwin junkie so I do this with a gawk script every day, but if you are not a scripting wizard you can do it the longhand way, by increasing the number of rows in the visible report to 100, sorting by descending CTR (click it twice), and then visually identifying the rows that have significant number of clicks. Then, take any domain which appears on this screen, and add it to your banned list.

3) If you are an engineer or product manager at Google, please, we could use some algorithmic help here. I realize this suggestion is going to cost you money in the shortrun, but when advertisers lose money you will eventually lose money too, because they will stop advertising. We give you all the information you need to calculate our maximum desirable cost per conversion (I have my doubts that we are intelligent in doing this, because you can use that information to screw us over royally, but business is based on a foundation of trust and for the moment I’m going to trust you). You should provide a setting (or make it default behavior!) that ads stop appearing on any site where they transparently won’t be profitable. I would also suggest screening sustained abnormal CTRs automatically for fraud or Terms of Service violations.

4) If you find a website which is abusive in their ad placement, you can complain to Google. Realistically, I think they value algorithmic solutions over manual ones so much that you have zero hope of being heard (and they have to — they got to being a gazillion dollar company by NOT having to pay a human to deal with the little shrimp with the $15 a day advertising budget). But if it makes you feel better, here is the link.

[Note: This post has been edited, as the author of the pictured site disputes my characterization of it, and claims that the effect was accidental. As I have no particular reason to disbelieve that, and his other pages do not appear to be exploitative, I’m giving him the benefit of the doubt and have edited this post to remove accusations directed at his site specifically. The technique, however, is being used by multiple sites and it strains credulity to think that eight people independently accidentally developed cross-browser compliant CSS and liquid layouts to achieve this effect.]

Major New Feature For AdCenter

Posted on September 6, 2007 by Patrick in AdWords

From an email I received earlier today, notifying me about upcoming changes to AdCenter on Friday:

Create campaigns with ease by quickly importing campaigns directly from other paid search programs or by copying existing ad groups within adCenter to a new campaign.

That is genius and honestly something I would never have expected from Microsoft, because it acknowledges that their competition is beating them soundly. That being said, the ability to import campaigns directly from Adwords literally halves the amount of work required, AND it means you get to do your work where the work is easy (AdWords’ vastly superior interface) and then reap the rewards of lower competition on AdCenter. That is a win-win-win.

Best Single Day Ever, Plus August 2007 Stats

Posted on September 1, 2007 by Patrick in AdWords, stats

Hideho everybody. I made some changes to my AdWords campaign right before leaving America, and lo-and-behold they increased the amount of clicks I was getting from about 10 a day to 60 a day, which contributed to me having my highest legitimate hits day ever (>500 visits, ~100 trial downloads). Some of that is probably due to the end of August being the start of the term, but after looking at the numbers it appears to be mostly AdWords. (100 trial downloads implies 2.5 sales implies about $70 in revenue, and believe me, if I could sustain that I would be sitting pretty.)

Anyhow, lets talk August numbers:

Sales: 33 (2 refunds, 9 CDs)

Income net of refunds: $813.45

Expenses:

Godaddy: $7

e-junkie: $5

CrazyEgg: $9

AdWords: $60.17

AdCenter: $11.28

SwiftCD: ~$65 (waiting on invoice)

Net Expenses: $157

Profit: $656

Lets see, what is of note this month:

#1 — Someone had shipping issues with their CD, and was a little miffed after it didn’t arrive, figuring it was my fault. It wasn’t, of course (I trust the SwiftCD sent it when their records says it did, which leaves it about 50-50 that it was customer error or the USPS up to its usual tricks), but of course I didn’t tell them that. What I did tell them was that I would UPS one out immediately (cost to me: probably close to $15), and I would have overnighted it if they had been any more upset. Yeah, that eats most of the profit from the sale, but they’re now very pleased with my responsiveness rather than thinking I’m a shiftless shipment-forgetting Internet conman.

#2 — I also sent someone a free CD rather than taking their school’s purchase order (PO). For those of you who have not sold to institutions, a PO is essentially a “Send us this stuff and then we’ll send you the exact amount of money on this document” transactional instrument. Dealing with them is a pain in the hindquarters — Quill had a whole DEPARTMENT of people whose only job it was to read school POs over the summer (one of them being me), and then there is another department for collecting payments on the ones that have been satisfied. I have only ever had one customer want to pay with one, and rather than spending hours of my life getting that mailed to me, then dealing with the school’s payment clerks to actually get my $29.95, I just sent them a free CD with my compliments. Now I’ve got a team full of teachers who love me and are hopefully plugging my software to parents, friends, and colleagues… who pay with credit cards, like normal people. :)

(P.S. If you’re in the position where you NEED to take POs, signing up with eSellerate or one of the other major shareware processors will work for you. Its one of the only times they earn their keep. However, since getting POs is as much a hassle for them as it is for you, be prepared to pay through the nose for it. eSellerate charged, if I recall correctly, $20 a month just to flag your account for accepting POs! I am skeptical that you can make the numbers work out very well on a $25 item. If you’ve got a $X00 item, though, get it done.)

Older posts