I remember years ago when I was talking with Mike Volpe and he said with a bit of frustration “Marketers are lazy.” This is a problem since training a machine learning model for any kind of marketing is a lot of labor.
Of course, you have to keep in mind that Mike, at that time CMO of Hubspot, was sitting in the eye of a perfect storm: he was the lead marketing executive running a team of marketers marketing a marketing tool to marketing professionals to do better marketing.
His frustration is real, many times software is bought in the hopes of making work go away. Tons of marketing projects are sold as “Hey! Let’s automate it! Yaayy, no more work!” That’s where the “lazy” comes from.
Where’s the work come from?
Unfortunately, new tools have implementation effort, and ongoing support (how many email sequences did you originally promise? How many are running? Yeah, maybe time to look at that again.)
Vendors don’t help. We’ve seen the cycle over and over again. A new category of tool is created. The vendors promise all the work will go away, marketers use that to sell it to management, the new tool comes in and guess what? It takes a bunch of work to keep it running.
All this with the wonderful irony that implementation of new marketing tools is a roll of the dice and the ongoing support is often downplayed to management. This would just be a normal day in any cube farm except for the fact that we’re talking about the marketing department, communication is supposed to be our thing.
There’s no easy answer to this, every project is an ongoing struggle between the outcome you want (and probability of getting there once started) and the resources needed to start the project (a.k.a. “Selling it to your boss.”)
There is a simple trick that you can use: Always get the “day in the life” story of implementation and support. Usually, you can’t trust vendors to give you this story so if you’re in the buying process it’s critical to spend some time with an existing customer to hear the true day in the life.
Rather than bore you to death with the example of training a machine learning model on some obscure topic like a subset of medical research, here’s the day in the life of my smart home project:
Project: Geek out my office so when I say “Computer, Red Alert” my office lights go red so I can feel like I’m Captain Kirk. The (granted, limited,) productivity gain is occasionally demonstrating nerd street cred.
The vendor says the story is “Set up the Alexa smart skill. Yaaay! You’re done!”
Here’s my actual “Day in the Life”:
- Change activation from “Alexa” to “Computer” because I want to be able to say “Computer, Red Alert.” I’ve created my first problem as the Alexa downstairs stays “Alexa” to appease the family and my inferior human brain now says “Alexa” half the time in my office.
- Create a command to go back to normal lights. I try “Computer, Green Alert” which fails, because that’s too close to “Green Light” which is obviously something Alexa gets asked about a lot because she really loves to tell me about the green light (Amazon has delivered my order, AKA porch pirate shields up.)
- Test “Computer, Close Bridge” which fails because Alexa thinks I mean “Close Garage” and then tells me I don’t have a smart garage door opener (which would be very bizarre not owning a garage.)
- I can’t use “Computer, Lights Out” because that’s too close to ‘Lights Off” which will shut off all the smart lights in the house, creating random yelling from rooms downstairs.
- Try “Computer, night bridge”
- Hey, it works!
Use day in the life as you scope your work, and stay out of the trap that’s caught all the lazy and their failed projects.
Footnote: For the gear geek contingent my semi-smart house is Alexa, Philips Hue light bulbs, and a Harmony remote control.
Need help with your marketing data and analytics?
You might also enjoy:
Get unique data, analysis, and perspectives on analytics, insights, machine learning, marketing, and AI in the weekly Trust Insights newsletter, Data in the Headlights. Subscribe now for free; new issues every Wednesday!
Want to learn more about data, analytics, and insights? Subscribe to In-Ear Insights, the Trust Insights podcast, with new 10-minute or less episodes every week.