In the run up to Measurement Week (September 15-19) everyone and their mother seems to be talking about measurement. Don’t get me wrong, I”m thrilled. But for me this my 1,698th Measurement Week, so welcome to the party. Nonetheless there are a few things I’ve learned along the way that I want to pass along before the party gets any noisier.

1,698 weeks ago, I did my first measurement project for Fujitsu Semiconductor. I’d been a journalist and an Asian Studies major and didn’t have a clue what I was doing. But I did a cost per lead and cost per impression analysis of competing semiconductor trade magazines, compared the results to the media coverage they’d given us vs. the competition, and then analyzed their audience reach data. As a result, I was able to carve $100,000 out of my budget and put it to better use.

My first benchmarking project followed shortly – I interviewed 20 of my peers in Silicon Valley to find out how much of their budget they typically spent on a product launch. – That got me a $3 million advertising and marketing budget for the following year.  I quickly learned that for a word person like me (working for engineers in Silicon Valley) nothing impressed quite like data, charts, and graphs.

1,569 weeks ago, shortly after I’d helped launch the first HP LaserJet printer, I was asked to do a promotion to computer dealers who had never before carried a Hewlett-Packard product. We put together a snazzy campaign and went a little over the $12,000 budget, so I was called on the carpet to justify the spend. I pulled a list of the dealers who received the promotional piece (none of whom had ordered from us before) compared it to the list of actual orders sold in the subsequent 60 days and calculated that our “over-budget campaign”  had generated $1.3 million in incremental revenue, or a 8567% return.

Fast forward another 218 weeks to 1988 when I worked for Lotus Development and realized that I was the 9th Director of Communications in my company’s 5-year history and that my boss’ boss hated me. I figured I’d better come up with some justification for my continued employment.

I thought about my goals and objectives and what we’d accomplished, then realized the one objective everyone agreed on was “to get our message across.”  I figured if I analyzed all our media coverage for the year to determine what percentage of it left someone more likely to do business with us, and what % contained one of our key messages, I might have a paycheck for another few months.  I took the 2,400 articles we’d generated, and had them read by someone who was in the market for software. The data showed that I’d increased message communications a steady 5-10% each month since I’d arrived, and had reversed a 3:1 negative to positive ratio, increasing the number of people exposed to stories that might leave them more likely to purchase their products by 60%.

When I presented the results, the head of my agency at the time, Bob Strayton at Hill & Knowlton declared that “any agency that isn’t doing this by the year 2000 doesn’t deserve to be in business.”  I quit and 1,341 weeks ago I founded The Delahaye Group and Hill & Knowlton became my first measurement client. Delahaye went on to pioneer most of the techniques that companies like Cision and Salience use today (probably because they bought my companies). It also produced some of the best Measurement Sherpas in history, but that’s another blog post…


I have alternately felt like John the Baptist “crying alone in the universe,” Cervantes tilting at AVE windmills, and Sisyphus pushing boulders up hill.

sisysphusSo it is very nice to know I now have company and it’s nice hearing all these voices out there talking so enthusiastically about measurement. While part of me longs to ask “What took you so long?,” another part remembers my entrepreneurial adage: “You’re never wrong you’re just early.”

Today, I feel like  Violet, the Dowager Duchess of Grantham — having seen it all and able to see through this nouveau niche of measurement.


So putting on my best upper class English accent, I bid you welcome to the party and now let me give you a few pieces of advice.

  • No one is that unique – and don’t believe anything you read on the internet. Everyone and their mother now boasts that they deliver ROI. No one really does.
  • Everyone gets their data from the same place. Most sentiment analyses comes from Lexalytics; most blog data comes from Boardreader; most metrics come from the same firehoses. The only difference is the account management and the interface, so ignore the bells and whistles and focus on what data you need to run your business/department/agency more efficiently.
  • Data is only good if it’s clean, and most of it is junk. Sorry, but it is the nature of my job to get a daily torrent of fictitious metrics, data that is full of stuff that shouldn’t be there, badly miscoded clips, and some of the strangest notions of measurement you can imagine.
  • Data without insight is just trivia and most people are puking up reams of trivia. Most of what your results will be ignored. This piece by Avinash Kaushik says it better than I can. No human can pay attention to more than about 5 things at a time, so if you have more than 10 metrics on your dashboard, 50% will be ignored.
  • No matter who it is, if they say they’re doing something new, chances are they aren’t. Been there, done that.
Recent Posts