ROI – Is statistical modeling the answer? Depends on the question.

15 Jun

Back in the saddle after much traveling and more than a few major project deadlines…  On May 23, PRWeek hosted a webinar with Dr. Hans Bender of Procter and Gamble, and Mark Weiner CEO of Delahaye, discussing P&G's statistical modeling efforts to determine the ROI of public relations.  First, here's a brief summary of the discussion:

  • P&G employed marketing mix modeling for six brands over a one to three year period.  Three of the six showed public relations with the highest ROI of any marketing tactic.
  • They employed multivariate analysis that relates changes in the marketing mix (advertising, direct, PR, etc.) to changes in business results.  The basic inputs were Impressions by week and sales by week.
  • P&G spent $5.9B in advertising during 2005 (OMFG!)
  • Factors other than quantity/impressions were used to factor PR impressions - primarily quality of coverage indicators like message delivery, type of article – mention/feature/exclusive, competitor mentions, and media type (online, TV, print, Radio)
  • The model allows for comparison of PR performance with other marketing tactics (advertising, direct, promotions, etc.), as well the ability to compare various elements/campaigns within PR 

So what did P&G find out about PR ROI?

  • PR drives sales, often on a par with advertising
  • PR delivers stellar ROI, much greater than advertising
  • PR provides a halo effect over other marketing tactics
  • Overall, P&G found about a 275% ROI for public relations
  • PR delivers high ROI on relatively low levels of spending

The last bullet raises an interesting question – to what degree is the ROI on PR scalable?  That is, can ROI continue at high levels with even higher levels of PR spend?  We would love to be involved with trying to answer this question with all of our clients!

In the promotion of the web event, the question was raised, Has P&G cracked the PR measurement code?  The short answer is No.  Not that the work they did isn't significant, interesting and highly relevant.  It's just that marketing mix modeling is hardly new, particularly with consumer companies.  We currently have multiple clients employing modeling that measures PR ROI.  What is new and significant about the P&G case is that they are on of the world's leading brands, and they were openly willing to share their experience and results. 

Marketing mix modeling is an area that deserves much more attention from the PR community.  It is interesting to note the story about the P&G modeling effort broke in the advertising, not PR, trades. 

The reason why marketing mix modeling is not THE answer to proving the value of public relations is that it really only addresses the media component of public relations.  One could argue about the percentage, but media relations is maybe 50% or a little more of what we do.  Brand PR, reputation management, public affairs, thought leadership, analyst relations, events, grass roots and crisis communications quickly come to mind as significant areas of PR involvement and importance where media coverage may not be the primary objective or measure of success.  Modeling also is limited in that it (mostly) addresses the sort-term, tangible impact of public relations.  PR, as we know, has both short and long-term impacts that are both tangible like sales and intangible like brand-building and reputation management.  Although marketing mix modeling is not THE answer to PR measurement, it is ONE of the answers we should be utilizing more often.  For some companies with large marketing investments, diverse marketing tactics, and visibility into end sales, it may be the best answer for demonstrating return on investment for public relations.   Plus, since PR almost always shows high levels of ROI and superior levels of ROI compared to most marketing tactics, it may be our best weapon to justify higher levels of PR investment. 

Thanks for reading.  -DB

13 Responses to “ROI – Is statistical modeling the answer? Depends on the question.”

  1. paul walker June 16, 2006 at 3:20 am #

    nice post, db.

  2. Mark Weiner June 17, 2006 at 2:12 am #

    Don,

    Thank you for listening to and covering the P&G/Delahaye webcast. You make excellent points, especially the one you make about PR being more than a marketing tool. While this is undeniable, there are statistical models that can help assess and predict the effects of PR in those areas where PR is also very effective: on corporate reputation, employee communication, public affairs, events, grass roots, etc.

    What’s being modeled in marketing mix modeling is the impact of PR (and other marketing agents) on sales. We’ve worked on models that assess the impact of PR (and other communication agents and external factors) on reputation, stock price, employee loyalty, and more. It’s just a question of the behavior your trying to understand and the extent to which data representing that behavior is available (stock price; employee turnover; voting records, etc.). As you know, I believe that statistical modeling of this type — whether confined to the marketing arena or extended to other areas related to PR and communication — is a seminal change in how PR is and will be practiced. It’s happening already: agencies who demonstrate sales win renewals…agencies who generate buzz lose. Special events that generate sales get investment; special events that don’t connect expenditure to sales are discontinued. If you work for a consumer products company or have one as a client, your world has changed irreversably, and you’d better get used to it.

    Luckily there are good sources available such as PR Week’s coverage of this webcast, which is available on the PR Week site. “Unleashing the Power of PR,” my book to which you refer later on your blog, provides details on P&G and several other marketing mix and PR-ROI cases.

    Keep up the fine work.

    Mark

  3. imi November 6, 2006 at 4:08 am #

    I found this page through search engine, when considering the measurement method for our next PR campaign. As far as in know, even if there is MMM in PG which pointed out high level of ROI of PR. PR still facing doubt on how many volume it would really generate by each investment on PR campaigns. PR is the only function fails to assume its incremental volume among all the integrated marketing actions. So, how can we actually make other people trust about the MMM result on PR from their heart?

  4. ph hanky January 10, 2007 at 2:00 am #

    who does P&G’s MMM work anyway? Delahaye isn’t a modeling agency.

    The problem with companies like P&G and Delahaye making sweeping statements like this is that there is no one to audit or verify the findings. Anytime you have someone doing a regression analysis like MMM, you have to question if the regression analysis was done with the right vigor and make sure enough factors were factored in the analysis.

    P&G has been making lots and lots of money and AG Laffley has been talking about their MMM and how they are cutting back on TV in favor of other vehicles. I don’t doubt P&G pulls off some world-class MMM, but I don’t know if they promote stuff like this to put the screws on their ad agencies and TV advertising rates.

    If they spent half of what they spend on TV on PR, then they would saturate the airwaves, the web, and print with PR. They don’t address diminishing returns on PR. I would assume you reach saturation point with PR long before you reach saturation on Tide commercials.

  5. Vipin Kandwal February 1, 2008 at 6:57 am #

    Hey all above said and done can someone suggest me any book which contains cases of PR-ROI measurability (if yes kindly forward the name and other details of the book to vipinkandwal@gmail.com)

    I also note that mostly the discussion is from the point of view of a company which uses PR but is not PR agency. More precisely, if I happen to be a professional in PR agency i would feel the way client sees/measures the ROI. No perspctive of an agency has been clearly talked about in the note or comments.

    I am a student, more information for analysis from experts would be highly appreciated.

  6. Don Bartholomew February 1, 2008 at 5:53 pm #

    Hi Vipin,

    Thanks for reading.

    Here are two sources that discuss ROI in public relations. First, here is an IPR paper authored by David Rockland, et al that covers basic ROI considerations and approaches:

    http://www.instituteforpr.org/research_single/perspectives_on_the_roi/

    You might also look at Mark Weiner’s new book, Unleashing The Power of PR. Chapter Seven – Real Business Results: Proving – and Improving – PR ROI.

    Hope this helps. I’ll drop you an email as well. -Don B

  7. Russ July 14, 2008 at 12:43 pm #

    DB,

    I am researching the use of econometrics in improving our evaluation of PR. What you have to say has made extremely interesting reading and it would be great to ask some further questions.

  8. Don Bartholomew July 14, 2008 at 2:53 pm #

    Hi Russ,
    Thanks for reading. I would be happy to discuss this further with you. Please send me an email (dbartholomew@mww.com) with a couple of proposed dates/times to talk.

    You may also want to talk with Mark Weiner (see his comment above). Mark can be reached at mark_f_weiner@gmail.com.

    Cheers, Don B

Trackbacks/Pingbacks

  1. Moderne-Unternehmenskommunikation.de » PR ROI - statistic modelling - June 23, 2006

    [...] Der “metricsman” beschreibt recht eindrucksvoll (”ROI – Is statistical modeling the answer? Depends on the question.“), wie Procter & Gamble den PR-ROI ber das “Marketing mix modelling” erhebt und welche Ergebnisse dabei herauskommen. Z.B. dass der PR ROI hher ist als bei anderen Instrumenten. Im Beitrag wird ausserdem darauf eingegangen, dass das Modelling seine Schwchen hat, da es vor allem den Medien-Output bewertet, aber nicht den Outflow wie Imageeffekte, Bekanntheit usw.. Aber immerhin wird hier PR nicht isoliert bewertet, sondern im Mix mit anderen Marketingkommunikationsmitteln. Fr mich ist die interessanteste Frage, warum das eigentlich in Deutschland kein Thema ist. Na ja, Thema schon, aber es wird nicht umgesetzt…. Oder irre ich mich? Verwandte Beitrge dazu: >>Texte zu PR Measurement und Evaluation [...]

  2. Online PR » Public Relations’ ROI - Promoting Goods and Services on the Internet - August 3, 2006

    [...] SOURCE: ROI – Is statistical modeling the answer? Depends on the question [...]

  3. Planet Relations » Archives » Procter & Gamble finds 275% ROI on PR - September 7, 2006

    [...] Here’s an interesting post from Don Bartholomew of MWW Group (a top ten PR firm) summarizing a PRWeek webinar on Procter & Gamble’s statistical modeling efforts to determine the ROI of public relations. The study examined the marketing mix (advertising, direct, PR, etc) for six brands over a one to three year period. According to Bartholomew, P&G used multivariate analysis that relates changes in the marketing mix to changes in business results. [...]

  4. Five PR Measurement Trends to Watch in 2008 « Proving the Value of Public Relations - January 16, 2008

    [...] and Gamble’s proactive outreach about their Marketing Mix Modeling (MMM) project (read more here) sparked a lot of industry debate in 2007 (and we suspect projects for modeling vendors).  The new [...]

  5. PR ROI – Measuring PR results to increase revenue | EASTWEST PUBLIC RELATIONS - June 5, 2010

    [...] ROI – Is statistical modeling the answer? Depends on the question. http://metricsman.wordpress.com/2006/06/15/roi-is-statistical-modeling-the-answer-depends-on-the-que… [...]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 171 other followers

%d bloggers like this: