Odlingsguiden, seven months later

In March I wrote about why and how I built a Swedish growing guide. Seven months later it’s time for an honest update.

Spring was intense. We went from 96 crops to over 140, added flowers as a fifth category and wrote the first longer guides. Then summer came and basically nothing happened. The season was in full swing and the farm took all the time we had. That’s just how it is when you grow things for real.

The honest problem

When we picked it up again this autumn we had to be honest with ourselves. Not as many people had found Odlingsguiden as we had hoped. Or thought, really.

It’s hard to reach out. We’re not salespeople and we don’t want to beat our chests about what we’ve built. We’d rather just make something good and hope people find it. Turns out hoping isn’t a strategy.

Fixing the basics first

The first thing I looked at was Google. And it turned out Google had only indexed a small fraction of our pages. Most of the guide was basically invisible.

Some of it was my own fault. Sections that were collapsed only rendered their content when you opened them, so Google saw an almost empty page. Navigation was buttons instead of links, so there was nothing to follow. That kind of thing. I’ll write about the technical parts in a separate post.

Search Console also showed what people actually search for. Not “tomato” but “when to sow tomatoes” and “how to grow tomatoes”. So we reordered the crop pages to answer when and how first.

Stop guessing, ask the data

Here’s the thing. We had a lot of ideas about what people wanted. But they were our ideas.

So I decided that before building anything big we should be able to answer whether anyone uses it. Every event we track in Odlingsguiden got a question attached to it. Do people choose their growing zone? Do they read the guides to the end? Do they use the search, and what do they search for that we don’t have?

To be able to answer those questions I had to build a lot of new stuff in Puls, our own analytics tool. Funnels, grouping events by data, markers for when a new version went out. Odlingsguiden pretty much drove the whole Puls roadmap in October. More about that in My growing guide wrote the roadmap for my analytics tool.

A new start page that didn’t get to be called 2.0

The start page showed vegetables. And that was the problem. Visitors saw the vegetables but never found the herbs, berries, fruit, flowers or the guides.

So the start page got rebuilt to show everything, with what’s happening in your zone this month at the top.

Internally the version number became 2.0.0, just to keep track of when we did it. But we decided not to call it “Odlingsguiden 2” anywhere. It’s not a new product. It’s a new look and more content. Same guide.

Min odling

This one came from Elin, one of our users. She wanted a timeline and calendar that fit her own growing. Not a lot of data about crops she isn’t growing anyway.

That actually changed something fundamental for us. From the start the idea was that Odlingsguiden is a reference work, not a planner. Now it’s a reference work at its core, with something personal on top. You add what you grow and get your own timeline.

It lives in your browser for now. No account, no login. If people use it, we’ll think about the next step.

Redrawing every illustration

Small detour. All the illustrations got redrawn in a new pen style, because the old ones more or less disappeared in dark mode. That made the pages way heavier than I’d like, and I had to rethink how the drawings were loaded to get the speed back. Another post for another day.

The first week

The new start page and Min odling went live in early October, and this is the first time we can actually see what people do.

It’s early. One week of data doesn’t tell you much. But some things were interesting:

• More people visited than the week before. Clearly more.
• A handful of people started adding their own plants to Min odling. Not many, but they added a lot each.
• Almost nobody chose a growing zone. Most people read everything for the default zone without ever changing it.
• Nobody used the feedback button. Not one.

The zone thing surprised me the most. The whole guide is built around zones. If people don’t choose one, they see dates that might be wrong for where they live. So that’s the next thing to fix.

There was also a spike one Saturday that didn’t show up as any source. A film about Lilla Bosgården, our farm, came out around then, so it might be that. Apps don’t send a referrer, so you never really know. That’s exactly why Puls got chart notes, so I can write “film came out” on a date and remember later.

What happens now

We wait and look. If people use Min odling, maybe accounts come next, so your garden follows you between devices. And maybe the app we’ve been talking about since the start.

If they don’t, then maybe not.

It’s not fun to build something nobody uses. And it’s better to find that out now than after another seven months.

If you grow things in Sweden, have a look. And if something is missing, the feedback button is right there. Nobody has used it yet, so you’d be the first.

The beds are put to rest for the winter now. Plenty of time to see what the data says.