Lots and lots of people use Google AdWords, and have more recently layered Google Analytics over the top. But the search giant has confused these users by re-defining a well-defined business metric: ‘ROI’. 

This has led to confusion within companies – and at worst, caused Google AdWords users to think their campaigns are more profitable than they are, and thus pump more money into Google while decreasing their own business’ profitability.

So has Google ‘done evil’ this time?

ROI – Return on Investment.  A term that has been around for a long time in the business world, helping you decide the best place to invest in your organisation. ROI analysis helps you answer questions such as:

  • Should you build a bigger factory for product A or B? 

  • Increase the sales or marketing spend for A or B? 

  • Sell A online only, or offline too?

  • Dump A altogether and invest in product C?

Wikipedia has a nice overview on the complexities of ROI: but it’s actually a simple concept. How does the extra money you make (the profit) compare to the money it costs you to get it?

It’s a ratio about profit – not revenue. Product A may sell at £100 a shot, and Product B at £200. But £1m in sales of Product A may generate more profit than the same £1m in sales of Product B if the ‘margin’ (sales revenue – cost of buying in and building the product) on A is higher.

Into this well understood use of ROI  – Google Analytics has thrown a spanner. It has invented a new definition for ROI. The result is that marketing folk are getting confused over that most fundamental of things - the incremental profit from incremental marketing activity. See this forum thread here on E-consultancy.

Google could have tried to help folks measure a true ROI – but that would have been hard - it would have to ask users to upload a matrix of their entire product listing, with price and % margin for each one. And to keep it updated each day.

That’s too hard, so Google instead decided it was enough to give users a metric to help them decide which AdWord spend gave the most sales revenue increase per £ of ad spend.

This is not ROI – more like ‘sales £ per mktg spend £’. If Google had called it that, no harm would have been done.

But calling it ROI is bad, for three very sound reasons….

Reason 1. It causes marketing types to look very silly when their business or financial directors find out that the lovely monthly reports showing excellent ROI were actually measuring something quite different.

And if your organisation has a seasonal sale model with sporadic high discount percentages, then everyone will get confused when ROI diverges even wider from the finance accounting metrics around that time.

Reason 2. The marketers in most eComm organisations I get involved with struggle to  control online business; in terms of overseeing how the technology is performing for users.

Firstly, they have a natural inclination to leave the technology to the tech teams; marketers didn’t study computer science at college. Secondly, when marketers *do* want to take control of the technology performance (in so far as it impacts the business), they can easily get rebuffed by the onslaught of jargon and non-answers the tech team provides.

Fundamental questions of importance to the marketer (and the business) don’t get meaningful answers, such as: ‘Did the extra traffic from my marketing campaign last week cause our online store to go slower, or throw more errors than normal, because I’m seeing lower sales conversion percentages despite the increased visitor numbers?’ (That’s the role of 24/7 User Journey monitoring, something I’m involved in daily with a range of organisations – that gives the answers and helps organisations decide how to spend the next marketing buck, or whether the online store infrastructure needs some investment first to increase sales conversions).

Using a non-standard definition of ROI increases the communication gap even with the tech team. Some of the senior tech staff will be studying MBAs in their own time or reading business books, and they’ll know what ROI really means, and their respect for the marketers isn’t helped when they uncover it’s misuse in the marketers’ monthly reports.

Reason 3. Trying to explain what Google’s ROI means is hard, because the percentage numbers it gives are not comparable business to business, or product range to product range.

How about this blog’s explanation: “Depending upon your business, you might need an ROI of 1,000% just to break even.” Clear as mud.

On the E-consultancy forum people are saying things like: “The ROI in your example would be =(5000-4000)/4000 = 25%.”  This sounds OK at first glance – like 25% profit. Coooool. But, actually these numbers really mean that the company spent £4,000, and got £5,000 in extra sales. Which means that unless its margins on those products sold is such that £,5000 of sales costs less then £1,000 to buy or build… then it lost money on that campaign. That is just not clear from the 25% figure.

Or people say: “ROI is calculated automatically in Google Analytics. I’ve never calculated any ROI figures myself since I don’t have a need to.” I’m not sure that all business managers would feel that such a blasé attitude to a key business metric is OK.

To conclude - it may be harder to get the product margin data from your financial systems or eComm reporting, but it is far better to use that for any ROIs you want to talk about within your organisation than to regurgitate Google’s fictions, expose yourself to ridicule and your company to blindly losing money on advertising campaigns.