roi

Having worked for a long time at an agency that resolutely refused to do any form of pre-testing I've had many long and involved conversations about measuring ROI. And they were always deeply frustrating and sterile conversations because we were never really measuring actual ROI, there was always some surrogate number in the way, and some embedded assumption about how communications is supposed to work. The ROI of communications is incredibly hard to determine. Even counter-productively hard.

So full marks to Alan Snitow who's trying to do some proper research on the topic. He describes it like this:

"The idea is to take a look and see what agencies today are actually doing, and how agency people really feel, when it comes to efforts to measure effectiveness and get to ROI for marketing communications.  For all the talk about it at conferences, by pundits, etc no one has ever systematically canvassed agencies to get a sense for what the lay of the land is.  I hope people will share their experiences so we can fix that.

And of course I will happily share my findings with anyone who participates."

Survey is here. Place to leave your email if you'd like a copy of the survey is here.

Comments

[Paul Fisher]

This is an incredible admission. Traditional advertising industry doesn't really know how successful their campaigns have been. Nor do they really try that hard to find out.

Performance Marketing: The online affiliate and lead gen model (CPAs and CPCs) has grown and thrived because its effectiveness is measurable. Exactly because it is ROIable.

As brands shift to the web, and these same brands get more and more success with performance marketing, what do you think is going to happen to traditional advertising?

Adspend will disappear. Traditional agencies will be dead.

Cheerio guys.

[Simon Foster]

A few comments:
1. Just done the survey; interesting questions. I find it hard to see how an agency answering "no" to too many of the specific ROI questions could survive long term.
2. The time when UK business efficiency was re-engineered from the post war period (80's), was also the last great age of non-ROI advertising. More than coincidentally, that was also the time when many agencies lost their seat at the "boardroom table." Companies got into measuring effectiveness and efficiency; most agencies did not match the pace of change.
3. The previous reader’s comment is too fatalistic. Yes, online is hugely powerful - I am one of its strongest advocates, and yes, advertising is in for an increasingly difficult time, but online activity needs advertising support – especially in highly competitive sectors. For example, anecdotal research is showing how advertising can influence search volumes for advertisers in financial services.
4. I believe there will always be room for a select number of agencies in the 'non-ROI' space. That's because some brands/owners know that really great brand positioning can be undermined by adherence to short-term ROI metrics. We wouldn't have had some of our greatest brand building advertising if ROI imperatives had ruled its development.
5. Being decoupled from ROI will be the ongoing reward for those creative agencies able to produce that rare advertising which is able to change the cultural status of brands.

[fran]

A very pedantic point, but important I think: Not only it is difficult to measure ROI, but even if we were able to do it, it would be a rubbish measure.

ROI is defined as:
(gain on investment-cost of investment)/cost of investment

basically it gives you a measure of how big your returns are compared to the advertising investment necessary to generate them.

It is a pretty useless measure because it doesn't allow to compare different investment projects (or advertising projects). For example:

- campaign A costs 50 and it will generate 100 returns. ROI=1

- campaign B costs 100 and it will generate 200 returns. ROI=1

Which campaign should you choose to run, based on ROI?

A much better measure is NPV (Net Present Value), which is the discounted value of future cashflows generated by the investment (or by the campaign).

This is getting even more pedantic, but my point is ROI is totally overestimated.

If you are really bothered , have a look at NPV here: http://en.wikipedia.org/wiki/Net_present_value