Google AdWords
Warning: this blog post may offend the short-sighted.
Our company has worked with a large number of online businesses over the years. And most understand the basic principles of how Google AdWords works.
We advise most of our AdWords clients to use conversion tracking, as this provides a useful gauge as to how well different parts of the account are working.
Note my wording. AdWords conversion tracking is a gauge; not that different from the RPM gauge on your car dashboard.
When you see that your RPMs are high, you don’t panic or celebrate. You simply factor that in to how and where you’re driving.
When you see that your RPMs are low, you don’t abandon your car, call the breakdown service or experiment with using milk as gasoline.
RPMs are measured on a scale. There’s a reason why your car doesn’t use a simple GOOD or BAD indicator for the purpose.
And like RPMs, conversion tracking data is useful. But it’s far from definitive, and far from accurate.
If the person buying your product is different from the person who clicked your ad, the conversion won’t register.
If the purchase takes place more than 30 days after the ad was clicked, the conversion won’t register.
If the PC used to order your product is a different one, the conversion won’t register.
I’m shocked by how many people I read about who slash their budgets or even close their AdWords accounts when they think it isn’t working for them. And guess which metric they use to reach that conclusion?
I’m shocked by the number of software developers who offer a 30 day trial, then wonder why conversion data doesn’t show up within five days of activating their campaigns.
“Most of our users buy within the first seven days” they say.
But AdWords visitors aren’t most of your users. They’re a new set of different users with different ideas and buying patterns. That’s why you’re using Google AdWords to reach them.
Don’t get me wrong. Questioning your account performance is vital, as is striving to improve ROI and efficiency.
But interpreting conversion data as rock-solid evidence of failure (or success) may be as sensible as cutting-off your head, arms and legs to cure a headache.
How do you differentiate between high clicks that are (or very likely) converting to sales and high clicks that are *not* (or not likely to) converting to sales?
Put another way, if I’m choosing between two approaches to boosting Adwords click throughs, and they are both costing about the same per-click, how do I decide which is performing better?
One thing to look at is non-sales activity (download, etc.) but even that can be misleading. If you oversold your product you might get folks who think you have the magic solution and come, poke around, download, ask lots of questions and then realize it’s over-hyped and never buy. They’d look like (and in fact were) very interested potential customers… but never bought. Likewise, if you were advertising to a demographic who wanted, but couldn’t afford your software (say you were selling ERP software and chose keywords that mom & pop shops were searching on: your software sounds great until they find out how much it costs).
Clay,
Unless you have high enough sales volumes you need to track download conversions. And IMO if your site is working correctly the download conversion trend should roughly match sales expectations.
So, those wrongly targeted individuals would see that what you provide is not what they were after and would not download anyway.
I think if you find you have an overwhelming increase in downloads with no following sales (after a sensible period of time) then it can’t just be the ad that is misleading, but the website too. The landing page probably needs work to clarify the offer.
Certainly if you had an ad that said “Magic solution to all problems, click here” and the page it landed on just said “Download Now” and nothing else then, of course, you’re going to get wasted downloads.
That’s an extreme example to illustrate my point. The ad is misleading but the website is even more so, so you’ll see more download conversions but no sales because it is not until they get to run the software that they find they’ve been miss-sold. If the landing page clarified the offer well then they’d not download, you’d see high clicks and no downloads and realise the ad isn’t working and change it.
And remember that Adwords helps you find a different kind of customer who thinks differently to your existing kind and to you, so as Dave says, patterns may be different and in the process of finding out what works and what doesn’t you have to go through some pain.