Ten B2B SaaS signups in ten days. Nine never told me what to do first.

I run a small practice with one job: finding where self-serve SaaS products lose new users between signup and first value. To keep the practice honest I sign up to a few products every week as an ordinary new user, record what happens, and send the recording to the founder. No pitch, no invoice.

Between 2 and 11 September I did ten. Sales engagement, review collection, popups, survey building, marketing dashboards, email marketing, content publishing, website building, visitor identification, digital asset management. Different categories, different price points, all self-serve on paper.

The same four things happened in most of them.

1. The first screen does not say what to do first.

Eight of the nine products that let me reach a dashboard landed me somewhere other than a single clear action. One dropped me on a feature page called “Sequences” with an empty state that said “no sequences yet” and nothing about what a sequence is. One put me on a seven task checklist with a 14 day countdown beside it. Two dropped me straight inside a builder, one for campaigns and one for templates, before I knew I had started building anything. One gave the largest block on the home screen to company news.

Every one of these products has a first action that proves the product works. Publish a post. Create a survey. Build a dashboard. Send a campaign. The product knows what it is. The first screen does not say it.

The one exception landed me on an empty home screen with a single line: create your first project. That is the whole fix.

2. The product asks me to leave before it has shown me anything.

Five of the ten sent me to my inbox to verify my email before I had seen one screen of the product. Four offered no social login. Two of those did both.

Verification exists to keep bots out. It does not need to happen at the front door. Let the user in, ask them to verify within seven days, and delete the account if they do not. The bot problem is solved, and the user has seen the product before deciding whether it is worth the trip to their inbox.

3. The questions are for someone inside the company, not for me.

Five of the ten asked for things a new user has no reason to give yet: a postal address, a required company phone number, my job title, how I heard about them, and in one case a five step buyer persona and ICP questionnaire before the dashboard.

The test I use: does the answer change the next screen? Two products asked what I wanted to achieve and then filtered the templates accordingly. That is a product question, and it earns its place. The rest asked, stored the answer, and showed me the same dashboard as everyone else. An onboarding survey that changes nothing teaches the user that their answers do not matter, at the moment you most need them to trust you.

4. Setup comes before value, and the clock is already running.

Four products put a step in front of value that needs credentials a new user may not hold: connect your Google account, connect LinkedIn, connect the Shopify store, connect your email infrastructure. One would not let me skip. One threw an error I could not read. Three had a 14 day trial counting down while that setup was still ahead of me.

At Hootsuite, new Analytics users landed on an empty product and had to connect accounts and wait for data before anything was worth looking at. We seeded new accounts with demo data. You could open a template, move things around, find a real insight, and understand what the product was for before connecting anything of your own. The aha moment stopped depending on setup.

What to do about it, in order of effort

1. Put one action above the fold on the first screen, written as an outcome. Nothing else competes with it. (A copy change and a layout change.)
2. Move email verification behind the first screen with a seven day grace period. (One backend rule.)
3. Add social login if you do not have it. (An afternoon.)
4. Cut every onboarding question whose answer does not change the next screen. Ask it later, once the user has done the one action. (A conversation with sales.)
5. Make every credential step skippable, and say what happens if it is skipped. (A product decision.)
6. Seed new accounts with sample data so the first action can be completed before any integration. (A real project. Also the one with the biggest effect.)

Where this is thin

Ten products is a pattern, not a study. I recorded as one user, once, on one device. I did not see any of these companies’ activation numbers, so I cannot say what any of this costs them. That is the question I would want answered before touching anything: which screen loses the most new users, and what does that cost per month.

If you run a self-serve product and cannot answer that, send me your signup link. I will record the first ten minutes and send you the video.