Landscape Mate

This is the sketch we started with. I made it based on Connor’s written and verbal description of a fuzzy idea he had while talking to ecological landscapers.

Cool! The design phase is done, right? I mean, the whole app is only 2 pages and an automated text message. How hard can that be to build?

As we often remind clients, the simplicity at this stage is a mirage. No matter how small an app sketch or mockup is, there are always more features than you think. And for each feature, there are a handful of technical and UX questions whose answers can conflict with one another and create additional features and complexity.

Let’s walk through the sketch. There are a surprising amount of features here.

First, the explicit features (the ones you can see):

1. Accept SMS from client
2. Accept MMS from client
3. View record (aggregate messages from all clients of a record)
4. View client message history
5. Send manual SMS to clients
6. Send automated SMS to clients once a month

Then, the implicit features:

1. Auth (sign up and login)
2. Create record
3. Edit record
4. Delete record
5. Create client
6. Edit client
7. Delete client
8. Associate record(s) with client(s)
9. Dissociate record(s) with client(s)
10. Manage LandscapeMate phone numbers

Now, for each of these 16 features, there are design (both technical and UX) questions that should be answered before we start building. Take, for example, the feature “send automated SMS to clients once a month.”

1. Do clients need to opt-in?
2. Can clients opt out? Can they opt back in?
3. If a client opts out, can they still send and receive manual messages with the phone number?
4. Is there an initial onboarding text?
   1. What should it say?
   2. Can it be edited? If so, what happens?
   3. Can it be turned off? If so, what happens?
5. What is the cadence (day, time, frequency) of automated reminders?
   1. What should it say?
   2. Can it be edited? If so, what happens?
   3. Can it be turned off? If so, what happens?
6. If a client is associated with multiple records, should they get multiple automated reminders or just one?
7. Should we automatically reply to client messages (something like “thanks, got it!”)?

Needless to say, our “tiny” sketch turned out to be bigger than we hoped.

So we did a collaborative design sesh in FigJam to explore ways to shrink the first feature set while still delivering enough value to get a couple of real users.

With a few strategic (albeit tough) decisions, we realized we could avoid a lot of complexity and save time during the build phase.

1. No organization account, just user accounts. If we get one person using this who wants to add their colleague, we will recommend they share a login and then build the feature into the app. That said, we architected the backend to be able to easily have this feature in the future.
2. Clients can only be associated with one record. This eliminates a lot of technical complexity. Plus, we aren’t sure landscapers even want to add multiple clients per record.
3. No ability to text clients back. This was the toughest tradeoff. We figured while yes, it will be awkward for landscapers to not be able to reply to messages they receive, the app will technically still do its job of automatically collecting site photos. Saving some build time to get feedback quicker is worth the risk.
4. No distinction between records and chats. Since there’s no ability to text clients back, there’s no need for an individual client view. All we need is a record that accepts MMS and SMS messages from clients.
5. No ability to edit automated messages or their cadence. We’ll explain that we configured messages to collect site photos regularly while not bothering clients too much. We expect this will be an issue for users, but we think we might be able to achieve our value without these features initially.

Now the question becomes: Did we choose the right features to eliminate and simplify? Are ecological landscapers going to feel comfortable using such a simple service? The only way to find out is to demo with potential users. If we discover that folks won’t sign up because it’s missing a feature we removed, that’s great! We can build it. But we believe iteratively adding features to a dead-simple foundation is a more effective (and faster) path to creating usable products than trying to push a big complex app in the right direction.

We knew we’d need to cold email folks in order to have great discovery calls. We just didn’t have a big enough network in sustainable landscaping going into this.

In the past when we’ve set out to cold email people, it never felt great.

As the studio sales guy, Connor was de-facto in charge of sending the emails. He’d find some email addresses, write a template, have one or two good days of sending, and then the process would start to break down. When no immediate responses came back, he’d feel self-conscious and start to overthink the template. He’d research each person more and more until he was sufficiently intimidated by their clout. He’d procrastinate and then feel guilty about not sending enough cold emails. It was a lot of emotional labor without a lot to show for it.

So this time around, we split the work.

I made a list of people we want to talk to and found their email addresses. I wrote the email template and customized the opening and closing lines for each person.

Connor then sent those pre-written emails to each person, managed follow-ups and scheduling, and conducted the calls.

The result? We sent 22 cold emails, leading to 12 calls with legit industry pros. Even though these numbers aren’t huge, they’re bigger than we've been able to do in the past. Plus, our goal was 5 discovery calls; we blew that out of the water!

The best part, though, is that we all felt better about the process.

I was able to develop a really ambitious contact list since I wasn’t nervous about emailing them directly. For the same reason, I also didn’t overthink the email template.

Connor was thrilled to spend less time on the initial email and instead put his effort where he has the most fun—responding and actually conducting the calls. Because the emails were written for him, he didn’t need to research people so he never got intimidated to press send. With a higher send volume, more responses came back quicker, which created a sense of momentum.

By splitting the work in this way, we removed the emotional labor from the process. That enabled us to send cold emails more consistently for a longer period of time. I think this is ultimately what yielded more calls.

First, we decided to focus on the ecological restoration/sustainable landscaping space. More on that another time.

Then we went looking for product ideas by having exploratory conversations with folks in the industry. It’s interesting how this Mom Test-esque philosophy—learn about people’s problems before building—differs from 12 startups in 12 months, a more product-first quantity-based strategy. I guess we’re taking a hybrid-ish approach, but I digress.

While we did talk to ~6 people from our existing network, we also wanted to be more intentional about learning from folks with a variety of roles and backgrounds. So we sent 22 cold emails, resulting in 12 conversations with popular authors, landscape designers, community organizers, grad students, and independent hobbyists.

We quickly learned that plant people are just, like, the best? Everyone was so kind and eager to help outsiders like us. One of the phone calls even resulted in a private tour of a green burial site outside of Philly, which in turn led to our attending the NDAL Symposium. Queue the brain blasts.

Connor conducted the “interviews” and recorded each one on Loom for reference later. I use quotes because there was no script or agenda other than genuinely connecting and learning about people’s problems. He prompted discussion with open-ended questions like “What’s the hardest part of your job?” and “Why aren’t more people doing these projects?” and “How do you think about technology?”

In classic Connor fashion, he gleaned many insights and ideas from these conversations (plus the conference and the books he was reading). He synthesized this qualitative data into a table of problems, potential solutions, and reasons for and against working on them.

Five ideas stuck out the most: two marketplaces, a social network, a fun design tool, and one SaaS-y product. I quickly sketched out each idea to help us pick one.

After a ~1hr discussion, we decided to try the SaaS product first. It’s the most likely to meet all of our criteria—must be in climate tech, use our product-making skills, and make sense for us to charge for Day 1. Right now, we don’t want to cold-start a marketplace and/or gamble on virality. Plus, even if this specific permutation of the SaaS idea doesn’t work, we’ll at least be talking to businesses with budgets. Hopefully, other similar ideas will arise.

So, what’s LandscapeMate?

Landscaping companies that do ecological restoration projects have a unique challenge—the sites need responsive management in order to ensure the intended outcomes are achieved. There’s no set schedule for when an invasive will show up or a drought will come or a species will fail to establish. Sustainable landscapers need to check their sites regularly in order to catch issues early, but they can’t afford to send crews around without a specific work plan.

Enter LandscapeMate—a tool that enables landscapers to automatically collect site photos. Twice a month, it texts site owners asking them to share a photo of the site. They simply text a photo back and LandscapeMate adds it to the site’s record, which landscapers can view from a desktop or tablet. With LandscapeMate, landscapers can remotely keep an eye on their sites over time in order to catch and solve issues early. With more successful projects comes more biodiversity. A win for people and the planet.

So the journey begins...