The ability for online advertisers to optimise their advertising is not new. Almost 100 years ago, Claude Hopkins, the father of Direct Marketing, outlined all the principles of ad optimisation – still relevant today – in his classic book, Scientific Advertising.
The gift of Pay Per Click advertising is that online marketers no longer need to wait weeks to count returned snail mail coupons – we can now optimise in real time.
In fact, the pendulum has actually swung the other way and for the most part, we are optimising far too early.
Not sure if you’re a premature optimiser?
Here are two signs to help:
1) Not enough data
Our PPC mentor and Adwords guru, Perry Marshall explains:
[bctt tweet=”Tests need at least 30 consistent responses for your numbers to be accurate within 5%”]
So no matter what you’re testing, typically ad creative, you need *at least* 30 units of data (eg. clicks/conversions) to have statistically sufficient data to even begin to make a comparison.
Making decisions where both the control and the test group don’t each have at least 30 units, is definitely premature optimisation.
Thankfully, the remedy in this case is rather simple… keep running the test until you have the required minimum 30 units.
(If in doubt, Perry Marshall has a Split Test tool to help you check your tests’ level of confidence).
2) Not enough difference
Without question, the bigger opportunity we see for *all* PPC advertisers is not in getting more data, but in having more vastly different creative to test.
If Claude Hopkins were alive today, he would have relished the opportunity not to major in the minors, but to learn the different appeals that work for different segments of a target market.
Most PPC platforms now allow us to segment (& therefore show different creative) to different age groups, genders, devices, customers vs non customers, new vs returners etc.
Its possible now to test dozens of vastly different creative for different segments within your target market – then optimise within each segment.
All of those good ideas that surface during ideation, that formerly would be canned for the ‘one best idea’ now have a chance to see the light of day – if they truly don’t work, we can always optimise them out later.
Going to your entire target market with the same few ads, is the most significant form of premature optimisation today.
As we edge closer to the ‘1 to 1’ marketing future, the biggest win for any PPC account are individualised big promises targeted specifically at select few audiences.
Once again, winners will be those who market ‘The Best for The Least for The Most‘.