• Home
• About me
• Who, Where, What

Can you measure the performance of brand advertising?

For the last 5/7 years performance advertising, primarily sponsored search, has been the clear winner of online advertising budgets. Banner hasn’t done bad, it just hasn’t done as good. Growth rates have been considerably lower in banner for the last few years, and quite a few are discussing the validity of banner advertising.

Usually, banner advertising is associated with brand advertising. This is a wrong assumption, as even banners have quite a big chunk of performance-driven budgets. And we are not talking about single digit percentage points, it’s double digits, and growing fast.

For simplicity, I will assume that performance advertising is exclusively driven by conversion rates. If a click converts into an action, then such click has value for the advertiser. How much is derived by the value the advertiser recognizes to the action triggered by the click.

Brand advertising does not rely on such immediate cause and effect relationship, and its value is better captured in terms of brand awareness, share of mind and the typical marketing jargon.

Performance advertising has been having a stellar performance because its ROI is fairly easy to measure -although not as easy as you might imagine-. Brand had a more difficult time: it wasn’t quite clear whether you could build, support or enhance brands online, while on the other side a conversion is a conversion. Probably the reason brand has been trailing performance is, amongst the many others about the novelty of the medium, that you cannot easily calculate the value associated with it.

Also, the two type of advertising are more interlinked and interdependent than you could imagine. Research shows that users are more likely to convert on a known brand, which means brand advertising reinforces performance advertising. Yet most of the value is credited to performance advertising, as it “closes” the loop. But this maybe can help deriving some value estimation for brand advertising indirectly from performance advertising.

A blind test would give interesting information: take two eshops, one of a famous brand, and another unknown, with all things equal (GUI, offers etc). Run performance advertising and check the ROI. Any difference, after having checked for statistical glitches, should be down to the value of the brand. This test would not capture the lifetime value of a user, so it would just give a hint of the value of brand, but I bet it is positive, and meaningful.

Clearly the value of the brand is not just driven by brand advertising. There is much more into a brand then a collection of billboards, magazine ads or TV ads. There is customer care, quality of products, desing, technology and many other factors. All these factors would be captured in that number too, hence biasing the result.

To partially fix this bias issue, you could run a blind test on two identical eshops, with one having run a massive brand campaign, and the other one no. Yet brands are developed over long periods, and not in just days or weeks. Still, I am convinced the “branded” website would achieve better conversion rates, and hence a less expensive performance campaign. This is how value of a brand campaign is transferred to a performance campaign, through higher conversion rates that result in lower costs. This difference would partially capture the value of brand advertising, and certainly put a floor to its rates.

Here’s a theoretical example.

Let’s assume we run blind test number 2, where two websites look exactly the same, but the second has a brand that has been advertised quite intensively in the weekds before. Both sites are targeting 5,000 customers and are running performance campaign at £3 per click. Unbranded site has a conversion rate of 2%, hence will need 250,000 clicks, for a total of £750,000 spend. Branded site has a conversion rate of 3.5%, therefore needing only 142,857 clicks for a spend of £428,571£. Not only, benefiting from brand recognition, branded website is able to get a £185 margin per customer vs £150, maybe thanks to upselling. Hence, branded site has a return of £925,000 vs £750,000 and could have paid a PPC of £6.48 vs £3.00. Basic mathematics allows to calculate the value of brand campaign at £496,429. Had the branded webiste spent less than half a million pounds, it would have got a better deal than unbranded website.

This is clearly a theoreitcal example, nevertheless the logic stands and the approach might help in better understanding the value of brand advertising, although many caveats apply.

Sphere: Related Content

Did you enjoy this post? Why not leave a comment below and continue the conversation, or subscribe to my feed and get articles like this delivered automatically to your feed reader.

Copyright Think-Through | Using the DeepBlue WordPress Theme

Bad Behavior has blocked 86 access attempts in the last 7 days.