Back when I was working at Next Jump, an e-commerce company that enabled big brands to offer their products and services at a discount to large employers and customers of large consumer marketers, our primary objective was to drive spend through our website.

My specific job was to drive user acquisition. I was focused on acquiring more companies to buy the product for their employees and then to get employees (users) to register an account and keep coming back. My colleague, I'll call her Jane, was in charge of site merchandising and had the job of converting those users into buyers once they came to our site. So my job was to get people to our site, and her job was to get people to buy once they arrived.

Every week our teams would meet to review results. We’d start by focusing on the total spend on our site during the previous week. Some weeks the numbers would be up and some weeks they'd be down. In the weekly meeting, our leadership would look at Jane and ask what happened during the previous week. Frequently, Jane would look at me and say, “we didn’t have a lot of spend on the site because we didn’t have a lot of traffic.” Other weeks I would look at Jane and say, "we had plenty of traffic but that traffic didn’t convert into spend."

This was obviously unproductive. We were pointing fingers at one another and defending our impact on the overall number which meant that nobody was responsible for the overall number.

Our solution to this problem might seem counterintuitive: we created silos.

We came up with something we called “the box.” My team had the job of getting people into the box (get people to the site) and Jane's team had the job of making good things happen once they were in the box (get people to buy things once they were on the site). My primary metric was weekly unique users and Jane’s primary metric was conversion of those users (spend per unique user).

This changed everything. We set up specific metrics for each team where neither one of us could ever blame the other. My team wasn’t measured on overall spend (something we couldn’t control alone) and Jane’s team wasn’t measured on overall spend (something her team couldn’t control alone). We were measured on our slice of the spend metric (users and conversions) and if we both did our job we had a great week. This change created crystal clear ownership and accountability which led to lots of creativity and powerful initiatives to drive each teams' numbers. Our overall spend numbers started heading up and to the right.

Over time, though, things started to break down. Because we were so silo’ed my team wasn’t focused on the overall company goal, we were focused on our team goal. So my team would do whatever we could to drive users to the site regardless of the impact on spend. We would repeatedly promote offers from Target and Best Buy (brands that had 'mass appeal’ and would drive traffic but had relatively low value discounts with low conversion rates). This would drive a ton of traffic to the site, but the traffic didn't convert. Similarly, Jane was focused on conversion so she would promote the best offers on the site (30% off Juicy Couture, as an example). Users would come to the site expecting to see an offer from Best Buy and would see a great offer from a brand they had no interest in and a not so great offer from Best Buy. This led to a low-quality experience, lower spend, and user churn. Overall growth in spend began to slow down.

In response, we quickly setup processes to begin working more closely together. We had to fix the disconnect. We had to collaborate.

We built a monthly merchandising calendar that every team member could access in real-time. We set up several 10-minute check-ins so that the acquisition team knew exactly what the site merchandising team was promoting each day and which offers were converting at the highest rates. The acquisition team would send all marketing emails to the merchandising team prior to sending to users to get their sign off. We used data from the acquisition team to convince the mass appeal brands to offer deeper discounts.

At first, these efforts forced collaboration. But over time the collaboration became much more organic. The teams became inclined to be collaborative. After a few weeks, the numbers started to head back up. That said, we definitely didn’t abandon the silo’ed metrics for each team. Hitting those metrics was still the primary job of each team. What changed was the approach we took to hitting each of our metrics. It was about transparency and collaboration and a broader focus on what was best for the company as a whole.

The point here is simple: not having silo’ed metrics is a bad thing and being too silo'ed is a bad thing.

As an example, sales teams need to have silo’ed sales metrics that they’re accountable for to force ownership and creativity and high performance. But if the sales team is only focused on one top line metric and nothing else, over time they’ll be motivated to close deals that may be bad for the company and will lead to high churn rates. They have to have a silo’ed metric but also be forced to consider what’s best for the company as a whole.

Companies get in trouble when they lean too far towards one side. Telling groups to just work together to drive an overall number leads to a lack of accountability and creativity. And too much separation leads to a lack of collaboration and focus on the broader goal.

Well run companies find a balance and learn to silo and un-silo.