Nobody Chose the $75 Column

Three companies. Same headcount. Same needs. Three different monthly bills.

The first one runs best of breed. Email and docs on the standard tier, a separate video meetings platform, a separate team chat tool, streaming on top. Fifty dollars per person per month.

The second one runs one suite, top tier, plus streaming. Sixty five dollars.

The third one runs the suite at mid tier and then buys video meetings, team chat and streaming separately anyway. Seventy five dollars.

Here is the part that matters. The first two are defensible. Best of breed can absolutely be the cheaper answer, and so can consolidating onto one suite. Both of those are decisions somebody made and can defend in a room.

Nobody chose the third one.

The third company got there because three different groups each solved their own problem, on their own budget, in their own quarter, and nobody was looking at the whole bill. There was no meeting where someone said let us pay the suite price and also buy the point tools. It just accumulated.

That is the entire subject of this post. Not which vendor. Not which architecture. The thing that costs money is the absence of a decision.

The money isn't growing. The tool count is.

Gartner forecasts unified communications spend growing at 2.4% a year through 2028. That is not a market expanding. That is a market holding roughly flat.

Meanwhile Deloitte's digital workplace research finds 90% of users run multiple platforms anyway, by their own choice, regardless of what IT standardized on.

Put those two facts next to each other. Spending is flat and tool count is climbing. That is not growth. That is the same money buying the same thing five times.

And nobody experiences it as waste, because no individual purchase was unreasonable. Every one of those tools got bought by someone with a real problem and a real budget line.

One group asks. Another buys. A third gets the ticket.

Zylo's 2026 SaaS Management Index, built on more than 40 million licenses and $75B in spend under management, puts a number on why this happens.

Business units control 81% of software spend. IT controls 15%.

Read that again with an operator's eye. The team that fields why did we buy this when five tools already do it has almost no say in whether it gets bought. They inherit the environment. They don't shape it.

The same research finds 36% of licenses go unused, and the average enterprise runs 305 applications.

Nobody in that chain is wrong. Marketing needed a thing. Sales needed a thing. The events team needed a thing. Every purchase was locally rational. But nobody is accountable for the total, and the invoice arrives regardless.

I wrote a longer piece about that ownership gap called Nobody Owns the Mic. This is the same argument with the bill attached.

Nothing new added. Just the increases.

Here is the part most people never model.

Take the $75 stack and the $65 stack and do nothing to either one. Add no tools. Change no vendors. Just let published list prices grow at 5% a year, which is the conservative end of what actually happens. Major suites raised list prices 5% to 8% in the most recent cycle alone.

What accumulated: $75 today. $96 in five years. $122 in ten.

What someone decided: $65 today. $83 in five years. $106 in ten.

The gap: $10, then $13, then $16. Per person, per month.

At 10,000 people that gap compounds to roughly $15 million over ten years, on today's stack alone.

Ten dollars a month per person does not look like a problem. It looks like a rounding error, which is exactly why it survives every budget review. It only becomes visible when someone multiplies it by headcount and then by a decade.

And all of that assumes nobody adds a single new tool. In twenty five years I have never once seen a company add nothing.

Where this goes, and why the clock is running

The reason to fix this now rather than next budget cycle is that the licensing model it depends on is about to come under pressure.

Gartner projects 40% of enterprise applications will run task specific AI agents by 2026, up from under 5%. By 2030 they put $234 billion of enterprise application software spend at risk as agents work across systems and bypass the traditional interface. By 2035 agentic AI is forecast at roughly $450 billion, around 30% of enterprise software revenue.

The mechanism matters more than the numbers. Per seat licensing assumes a human sitting in a seat, clicking through a product. Agents don't need seats. When work gets done by something that doesn't log in, the link between headcount and revenue breaks for the vendor, and every vendor is going to reprice to protect it.

If you are paying five times for the same seat today, a repricing is when you find out.

This was never a vendor problem

I have consolidated platforms at enterprise scale and I have sold into companies doing it badly. The pattern is the same everywhere, and it is not a technology failure.

Either lane works. Best of breed can be the right call. One suite can be the right call. The expensive column is the one where nobody made a call at all.

Somebody has to be willing to own the whole bill. Not the tool. Not the contract. The bill.

Until someone is, the third column keeps growing at 5% a year, quietly, on a line item nobody reads.

---

All prices above are published 2026 list prices from Microsoft, Google, Zoom and Slack, annual billing, rounded to the dollar. Streaming is held constant at $5 across all three paths as an illustrative figure, so it changes none of the comparisons.

Sources: Gartner, Forecast: Unified Communications, Worldwide. Gartner, agentic AI enterprise software forecasts 2026 through 2035. Deloitte digital workplace research. Zylo, 2026 SaaS Management Index.

For 25 years I've helped people and companies communicate using AI, cloud and video. If you're staring at a stack nobody owns, I'm always open to the conversation.