Arlo Mistake #5 - Avoiding the Hard Integrations
As I outlined in my first mistake post, Arlo started as an internal tool and expanded from there. When we used it internally it had simple integrations with our Customer Relationship Management software (CRM) and Slack for notifications. This allowed our team to stay on top of changes.
When we took the product to market we did not integrate with any other systems and it was the most common request. Do you integrate with our CRM (Salesforce was highest and HubSpot second)? Do you integrate with our data room system (Box, Microsoft OneDrive, DropBox, Google Drive, etc)? Do you integrate with our email (mostly Microsoft 365 but some Google Workspace)?
The answer was "no but we are considering adding them to the product." I knew that customers would want to integrate between Arlo and their own systems. I knew it! I also knew that any one of these integrations was a big lift... not just the technical integration itself but the workflows within it. Customers may not realize it but they have hidden expectations around how something should work especially when mapping against their own personal workflows with the exact same tools.
I pushed off the problem by saying that we have an API and you are welcome to integrate Arlo into your own systems in whatever way you want. Goodness. Even just writing that sentence now makes me shake my head. While saying that we had an API was easy for us, it pushed all of the effort onto the customer. (Side note: we did not have a public API but if someone asked for it we had a plan to pull one together in a day or two). No real surprise that over the year of building Arlo not a single company asked for the actual API specification. No one cared to build out integrations for a tool they weren't sure they were going to use or adopt at that stage in the sales cycle.
Looking back, I avoided these integrations because I was overwhelmed by their complexity and wanted to focus on what seemed like easier wins. Each integration meant not just connecting systems, but understanding and supporting dozens of different workflows. But that calculation was wrong - the "easier wins" weren't wins at all if customers couldn't fit Arlo into their existing workflows.
Eventually we did integrate into the customer's email system and that became a very valuable feature that continues to offer value. The integration allowed Arlo to automatically send reminders to companies applying for financing that documents were still missing - a huge pain point for sales.
If I had to do it over again I would have integrated with Salesforce (top CRM requested), Microsoft Office 365 (top email requested) and on the data room side I would have integrated with Box and Microsoft OneDrive (two most common requests). We knew about the prevalence of these tools early on and I avoided the hard integrations that would have driven significantly more value simply because they were hard.
The real lesson here isn't just to "lean into the hard things first" but rather to truly look at how you can provide value. For B2B software, integrations aren't just features but rather they bring your software into the workflows of your customers. Start by identifying the top 3 systems that show up in every sales call, build deep integrations with those platforms and validate with early customers that you are nailing the workflows.
I'm not the only one to make this mistake and I won't be the last.