The Journey Map That Told Us to Do Less

In my years at General Motors in the mid 2010's, my team got assigned a deceptively simple question: why weren't customers signing up for OnStar when they bought a new car?

We did the whole thing. Months of research: dealership visits, customer and salesperson interviews, a journey map covering every twist from first awareness to driving off the lot.

The map told us something strange. Customers knew OnStar. They knew the features. Awareness was fine. Somewhere between awareness and the purchase process, things fell apart.

So we did what teams do: we added more. Better advertising. Cleaner education materials. Fixed every piece of online and in-dealer content we could. And we kept drilling until we hit the one place we were told was off limits: the dealership process itself.

That was the aha moment, and it was ugly.

On average, buying a car took about three hours from sitting down to driving out. And we were going to ask dealers to add OnStar education steps on top of that? No dealer wanted that.

Then we looked at what filled those three hours, and found the real problem. The customer's data was captured once for the brand, like Chevrolet or Cadillac. Then again for the dealership's financing. Then a third time, on an entirely different set of forms, for OnStar. Salespeople were manually re-entering the same information three times because three different parts of GM each required it.

And behind each capture sat a different incentive. The brand side said: get them through fast, the test drive sells. Financing said: slow down, they need to understand payment plans. OnStar said: if you don't tell them about the service, we can't turn it on, because it requires consent. Three divisions pushing and pulling on one person in a dealership chair, and the customer just wanted to buy a car and leave.

Every part of the process had an owner. The moment they were all fighting over had none.

So we restructured around the customer instead of the divisions. Pull the data from where it was already collected. Ask the bare minimum. Move everything that could move to post-sale, when the customer is pleased and actually curious about what they just bought: a text, an email, help with the mobile app. Equipped, but not inundated.

The result, in round numbers, since that's what I trust my memory to: OnStar uptake at dealerships went from roughly 20% or below to north of 90%.

No new features. No new ad budget. No education push. The breakthrough was deleting work: three captures of the same data became one, and a lecture became a follow-up.

When a metric is stuck and everyone proposes doing more, the actual blocker is usually something no one owns: a moment in the process that every team touches and no team is responsible for. You fix it by finding the duplicate work and taking it out.

Where in your product journey is the same data being captured twice, because two teams each need it for their own system?