Tuesday, February 5, 2013

Your success metrics aren't unique (or, why are we making analytics so hard?)

Last week I had the pleasure of attending TechKnowledge 2013, including giving a workshop on game design with Karl Kapp and presenting a concurrent session on immersive design. Two action-packed days in San Jose, for sure, and for once I got to see some of my esteemed colleagues' presentations as well.

I was most excited to see Ellen Wagner present on learning analytics and get a sneak peek into some of the work that she's been doing with the Gates Foundation. I was not disappointed; I loved hearing Ellen's practical advice to the packed room on how to approach analytics and not be intimidated by big data and the necessity of quantifying learning...or improvement, really...for our non-learning business colleagues. I appreciated her encouraging us to be able to frame the value that we bring to the organization in business terms without abandoning our expertise that qualifies us to take on an important organizational leadership role.

Watching Ellen's presentation left me thinking (which is normally a very good or very dangerous thing).

Every presentation that I've seen on analytics and big data end with the same call to action: "now, go back to your organization and figure out what you should be measuring."

Photo by WH "Snowflake" Bentley
Here's my challenge to you, my esteemed learning colleagues: shouldn't we all be measuring (basically) the same thing? We all suffer from the snowflake syndrome (I'm just like everyone else! I'm unique!), but to be fair, both statements are true. The real question is how much are we alike and how much are we different?

About ten years ago, I was working with a gentleman who had done an in-depth analysis across 15+ pharmaceutical companies, looking at what made their sales representatives successful. He found that 80% of the success indicators were exactly the same across all companies. The unique 20% related to the type of product, the therapeutic market the drug was being sold into, and the corporate culture differences among the companies.

Now, if you're a pharmaceutical company, where do you spend your analysis energy, knowing this? Right. Don't worry about identifying success markers for the 80%: use the identified metrics, provide training and support for those areas, and spend your analysis time identifying what the other 20% are that are unique to your company. EVERYONE should be measuring that 80%...that's baseline. Figuring out the 20% is what differentiates you from your competitors and supports your brand.

Could the same analysis be achieved for call center support? Project managers? Database engineers? Are there consistent success metrics that are common across industries?

If we could identify the success metrics across within or across industries, we wouldn't have to start from scratch to determine our analytics. We SHOULDN'T have to start from scratch. Don't we know what these things are? Can't we say, "Hey, you want to measure your organizations' success and performance improvement? Here is 80% of what you should be looking at, and what it means. Now, figuring out the 20% that's unique to your organization is up to you." (I'm not saying that it's literally 80% for every job function or every industry...but you get the idea.)

As learning professionals, we should know these benchmarks for our industries. If we don't, that is our very first starting point: find the "80%." What makes us the same? What makes us unique? Identify it, benchmark it, measure it. Then? Spend your time analyzing what makes you better than that 80%.


  1. Koreen, this is exactly what my company is attempting to solve. Our CTO recently said "the eLearning world is more self-similar than it is different," in the context of learning-related data analysis and visualization. I couldn't agree more that many of the success measure and ways we analyze data are (or should be) fairly standardized.

    Part of the problem in the L&D industry is that we are inundated with consultants who are telling us to go "figure out what we need to measure." The other part are vendors who don't know how to build/provide measurement tools, and simply deny the fact that it can be done - stressing the need for custom reporting and measurement tools.

    1. I'm going to do a session at DevLearn this year on this 80:20 idea, and what I think falls into the "80" category...will be thinking about this a lot more, and would love to get your feedback/input!

  2. This comment has been removed by the author.

  3. Thanks for this post.
    we are expert in effective branding marketing you can contact
    feel free to Rich Media thanks . . .

    marketing advertising in singapore