We signed up for Zendesk in 2011. They were a startup then and so were we. I don't remember what we paid that first year, but I remember it felt fair. Fifteen years later we are going to sign what will be our final renewal with Zendesk, and I want to explain why, because I suspect a lot of companies are sitting on renewal quotes that look exactly like ours.
Quick background for anyone who doesn't know DoiT. We're a 750-person company that helps businesses get more out of AWS, Google Cloud, and Azure. Our platform (Cloud Intelligence™) and our forward deployed engineers manage billions in cloud spend, and most of what we do, one way or another, comes down to making our customers' bills smaller. The irony of what follows was not lost on anyone here.
What happened
Over the past couple of years our support team automated more of the workflow, improved self-service, and got fewer tickets in front of humans. By renewal time we needed 175 Zendesk seats instead of the 279 we had been paying for.
We went into the negotiation expecting to save around $175k. Cut a third of your seats, save roughly a third of the money, or so we assumed.
The first quote came back at $413k, which is what we paid last year. They had raised the per-seat price by enough to absorb the entire reduction.
A hundred and four fewer seats, same bill, and if you do the math on it, we would have been renewing at essentially list price after fifteen years as a customer.
We pushed back, obviously. The revised offer was $372k for 180 seats, so $41k in savings, plus a 3.5x bump to our API quota, which we did want. They also offered $354k a year if we signed for two years. We declined that one, since committing two more years to a vendor mid-breakup made no sense to us.
Add it up and we reduced our usage by about 35 percent while our bill went down about 10%. Somewhere between the $175k we planned to save and the $41k we were offered, the rest of our savings became their revenue.
The part that bothers me
I want to be careful here because I'm not writing this to complain that software is expensive. We sell software. Our customers pay us real money and I think they get more back. Price wasn't the problem.
The problem is what the price was doing. When we cut seats, they raised the rate, and the only purpose of raising the rate was to make sure our efficiency cost them nothing. The work was ours and the savings were theirs. The whole maneuver depends on a bet that migrating off a support platform hurts more than overpaying for it, and for most companies, most years, that bet pays off.
It turns out procurement people have a word for this. When you reduce scope at renewal it's called descoping, and unit-price increases in response are a standard vendor play. Vendr's benchmark data has Zendesk customers describing our exact experience, including one buyer whose unit rate went up 15 percent for descoping and another who wrote that when they decreased licenses, the price "went up significantly." Vertice's SaaS Inflation Index puts SaaS inflation above 13 percent, several times general inflation, and found more than a quarter of contracts affected by some form of shrinkflation. Zendesk didn't invent any of this. They ran the industry playbook, and we happened to be the customer it ran on this quarter.
I think there's a reason it's getting worse right now. Seat-based pricing made sense when headcount only went up. AI ended that. Support teams everywhere are handling more volume with fewer people, seat counts are falling across the industry, and vendors who charge by the seat are watching their contracts shrink. Some of them will change their pricing model. The rest are defending last year's number by charging more per seat, which amounts to billing customers for their own efficiency gains.
Why we can't just shrug this off
When a DoiT customer cuts their cloud bill by 35 percent, we send congratulations. Sometimes we're the ones who found the savings, and the customer's spend through us drops as a result.
We built the entire company on the idea that this is fine, that customers who save money stay longer and tell their friends, and that eating a smaller number today is how you earn a bigger relationship over time.
So when a vendor of ours does the exact opposite, it grates. Accepting it quietly would mean our own pitch is naive. I don't believe it is, but I'm also not going to keep funding the other model with a straight face.
The next twelve months
We are going to renew for one last year.
There's a saying I'm fond of: it's not expensive if it's for the last time.
During that year we'll either move to a vendor who prices honestly, or we'll build the thing ourselves.
If we find a vendor we like instead, great. Commercial or open source, we don't care much, as long as the pricing gets cheaper when we get better instead of the other way around.
If your (Zendesk) renewal looks like ours
Start earlier than you think you need to. Bring your usage data to the table. And decide, before the negotiation starts, whether you're willing to leave, because that willingness is the only leverage vendors like this respect. For fifteen years we weren't willing. Now we are, and it's remarkable how different the conversation feels.
I'll write a follow-up in a few months with what we picked, what it cost, and what the migration was like. If you're going through the same thing, my inbox is open, because judging by the procurement data there are a lot of us.