Automattic: what WordPress.com looks like to someone who has never used it

The short version

Automattic makes WordPress.com. In the summer of 2025 we were already talking about working together on something else, and I pitched them a separate engagement first: pay me to go through WordPress.com the way a new customer would, because the people who build a product every day are the last people who can see it fresh. They said yes.

I had built websites on WordPress since 2017 and had never used the .com side. That was the whole point. I signed up, got confused in the places a normal person gets confused, and said so on the recording while it was happening.

Why they wanted an outsider

WordPress.com has a problem that almost nobody outside the WordPress world understands. There are two things called WordPress. One is the free software you install on your own hosting, and the other is the hosted product Automattic sells. Every regular person I have ever talked to just hears “WordPress” and has no idea there is a difference. Plenty of people inside the ecosystem are fuzzy on it too.

So the product has to do two jobs at once. It has to catch a total beginner who wants to start a blog, and it has to convince a freelancer or an agency owner who already has strong opinions about WordPress. I sit in the second group and I talk to the first group constantly, which is why they asked.

How I did it

No script and no guided tour. I went to the homepage and did what anyone would do.

I clicked the main button, created an account, went through the domain step, built a site on the free plan, launched it, and then upgraded to the Business plan to see what an agency owner gets. I tried the AI site builder. I dug through themes, plugins, the dashboard, and the settings. I narrated the whole thing as I went, including the moments where I had no idea what I was looking at, because those moments are the actual findings.

The recording ran about an hour and a half. I paired it with a written report, organized so their team could jump straight to the section they owned.

What I found

The product itself was capable. The problems were almost all about what a person understands at each step.

The first button never said free. The homepage said “Get started,” and nothing told you whether you were about to pay. If you are about to connect your Google account to something, you want to know what happens next. Meanwhile the pricing page did not show the free plan at all, so a beginner had no way to connect “Get started” with “this costs nothing.”

Checkout defaulted to three-year pricing. When I went to pay, the monthly price I was looking at was the three-year rate, and it was not labeled anywhere obvious. I said on the recording that nothing else in your life besides a car lease comes in three-year terms. Nobody expects it, and a price that turns out to be something other than what you thought reads as a trick even when it is not meant as one.

Signup pushed a domain on people who might not know what a domain is. The option to skip it existed, but you had to go looking. Someone who just wants to try it before spending money should be able to get in, play, and hit a clean upgrade path later.

Two audiences, two different frustrations. This was the most useful frame to come out of it, so I will quote myself: “The lay people are going to get annoyed by being completely lost and not knowing what’s going on. The agency owners are going to get mad because there’s too many things in their way.” Beginners needed education up front, and the free courses were buried in a dashboard carousel. Agencies needed a real home for managing a stack of client sites, clearer collaborator roles, and fewer hoops. One fix does not serve both.

Plugins were a wall. At the time you could only install plugins on the Business plan and up, and the wording flipped between “install” and “activate” in ways that made it hard to tell what you could actually do on your plan.

The AI builder was fast and shallow. It made a first draft quickly, but it asked too few questions to build anything meaningful, the images were generic, and a site built with it could not be launched on the free plan even though a site built by hand could. That last one I only discovered by trying to launch it.

The part that was hardest to say

The last thing I said on the recording was the most important thing in it.

I told them I could honestly say WordPress.com was powerful. I could not honestly say it was easy. The help was there, but not in the places people get stuck, and until that changed, anyone telling a beginner “this will be simple” would be setting them up to feel stupid.

That was not a fun thing to tell the company paying me, but an outside read that comes back as a list of compliments is not worth what they paid for it.

What happened after

Their team lead watched the whole thing, all hour and a half of it, and passed it around internally. A couple of weeks later they asked me to scope more of the same kind of work.

About a month after I handed it over, WordPress.com started letting the Personal and Premium plans install plugins. It began as a two-week trial and became permanent that September. Today their pricing page shows the free plan first and defaults to yearly billing.

I cannot take full credit for that. Their own announcement said a lot of people had been asking for plugins on the cheaper plans. But all of it was on my list, and I would like to think I had some influence on how it turned out.

The relationship kept going too. Before WordCamp US in Portland that August, they asked how I would run the recording studio they were setting up for creators at the event, and I sent back how long sessions should be, how people should book them, and why lighting mattered more than cameras. After the event I recorded a second round of feedback on how it actually went. We also talked a few more times after that about other feedback and projects.

What I would do differently

I would agree up front on how we would know if any of it mattered. I handed over a thorough read and never asked which findings they planned to act on or what number they would watch. So a year later I can point to changes that line up with my list, but I cannot prove anything, and that is on me, not them.

I would also put the debrief on the calendar before handing anything over. I sent a booking link with the deliverables, and in a company that size a link is easy to lose, so the call where we would have talked through what to fix first never happened. The recording and the report did their job, but the conversation is where feedback turns into a plan.