This is part six of a six part series about measuring return on investment of public relations.

Econometrics 

Even in a complex example of an outcome being affected by many factors there are tools that can help us.  These generally come under the heading of ‘econometrics’.  Econometrics techniques have been around for a while - as a discipline it was founded in the late twenties by Nobel prize winning economists Ragnar Frish and Jan Tinbergen.  The bit of econometrics that is relevant to ROI is known as market mix modelling which uses statistical techniques to disaggregate the various effects including splitting out the influence from different forms of marketing.

By doing this it is possible to work out the percentage contribution of a marketing activity to an outcome, say sales, and therefore to a financial return – the magic R.  If we know the cost of each activity, the I, then we can establish the comparative ROI

Econometrics has proven to be a very popular technique to prove ROI for advertising.  It has been notoriously hard to prove that advertising works ever since Philadelphia retailer John Wanamaker said “half the money I spend on advertising is wasted; the trouble is I don’t know which half” back in the 1860s.  Econometrics has gone someway to answering this question which is why many advertising agencies back up their work with econometrics and a number of larger advertising groups own their own econometrics companies.

This is both an opportunity and a threat to PR.  The opportunity is that many organisations are already developing econometric models.  The threat is that it is often driven by advertising or broader marketing and PR is often not included.  The problem with this is that models are very sensitive to the data that they are given – if there is no PR data then none of the outcome is attributed to PR.  Even if that hard won editorial coverage may well have influenced your audience to act, the model will attribute all of that to those activities that are included.

As an illustration of this Metrica had a client recently whose advertising planning company had built an econometric model without including PR.  Not surprisingly none of the outcome (in this case web traffic) was attributable to PR with advertising taking claim for a significant proportion – 18%.  We rebuilt the model with PR included.  This time the model showed that 8% came from PR with just 10% from advertising.  In the earlier model the traffic that should have been attributed to PR had been attributed to advertising.

The moral of this story is that PR practitioners should act to find out what modelling is already underway in their organisations and make sure that their work is included.

Incidentally in the example mentioned, web traffic was in turn directly related to income.  Because of this it was possible to calculate that PR’s 8% contribution was responsible for delivering £5m.  The PR budget – nowhere near that.  How’s that for return on investment!

Over the past six posts we have looked at a number of techniques to measure ROI.  While AVEs continue to be by far the most popular within the industry we have shown that there are a number of alternatives from the relatively simple (cost of reach) to the more advanced (correlation and econometrics).  All of these alternatives are commonly used by other marketing disciplines, so surely it makes sense that if PR is going to compete at the top table that it uses a similar approach.

Previous: Part 5 – Correlation with Business Outcomes