Facebook has been putting a lot of effort into growing Messenger as a bot platform this year, and now there are 34,000 of them in existence, built to automatically give you news and entertainment, let you shop, and more, expanding Messenger’s use beyond simple chats with friends. Today, that strategy is getting a significant boost: Facebook says it will now make bots trackable on its free analytics platform, alongside analytics for ads and apps. And Facebook is also opening up its developer program, FbStart, to bot developers as well.

Both potentially give bot makers more reasons to build and monitor how their new widgets are working.

Josh Twist, a product manager for Facebook’s bots efforts in Messenger, tells me that Facebook expanded the analytics and FbStart tools after a lot of requests from the developers.

“Getting bot support for messenger is the most frequently requested feature,” he said. This shouldn’t be too much of a surprise: Facebook already provided these kinds of tools to other developers on its platform, and bots have seen a huge surge of interest, both from users interested in trying these out to see how they work and also developers keen to see if this is the next big thing.

Analytics, of course, is an essential tool for a developer, both to be able to track how well something is working and other kids of feedback. Here Facebook says that features that will be included are reaches across mobile and desktop devices and measurement of customers’ journeys across apps and websites.

Developers also will be able to view reports on messages sent, messages received, and people who block or unblock your app. And they will also get access to anonymized data reports on bot demographics, which include details like age, gender, education, interests, country and language to figure out who is using your bot.

FbStart, meanwhile, currently has some 9,000 members who get feedback from Facebook on their apps, ads and bots, as well as Facebook ads credits and other free tools from partners like Amazon, Dropbox, and Stripe. If Facebook was looking at ways of swelling those ranks, tapping 34,000 developers could be one way of doing that.

Twist points out that while there are a lot of standalone bot developers coming to Facebook for the first time, there is a lot of crossover with other Facebook services like apps and ads. Those who are leveraging these together — for example using the recent ability to channel a person from a News Feed ad through to your Messenger experience — will be able to look at the effectiveness of those efforts now, and make potentially more ad buys based on them.

Twist tells me that for now, the analytics will cover bots built just for Messenger, although don’t be surprised if Facebook expands it to other platforms. “It is something we have talked about and haven’t ruled it out,” he said. “It’s possible, absolutely, since we already support analytics for other platforms for apps. But right now we’re prioritizing support for Messenger bots.“

For years, a truism in software investing was that the value of application software lies in data, not in technology. Companies like Salesforce, Workday, and ServiceNow are valuable because they are the “system of record” (SoR), or single source of truth, for their customers’ most valuable information, such as customer records or employee data.

As a result, they become deeply embedded in their customers’ business processes, making them hard to rip out. That gives them tremendous revenue predictability and pricing power. The technology itself — databases combined with workflow engines — is not particularly innovative; it’s the information captured by the technology that’s important.

The newest crop of software applications turns this logic on its head. They mimic consumer companies by using technology as a “wedge” to gain widespread adoption and don’t even try to become systems of record.

Instead, they are “systems of engagement” (SoE), meaning apps that employees actually use to get their work done. For example, take Slack, which Forbes recently identified as the most valuable private cloud company.

The data in Slack is either low value (“water-cooler” conversations) or already lives in existing systems of record. The same is true for many other fast-growing apps, like Intercom (customer interaction), Clari (sales), Culture Amp (employee feedback) and Front (shared inbox).

Digging deeper, the specific areas of technology where these companies have innovated are ones that historically people have ignored — integration and design.

At big companies, integration is the ugly step-child of any product roadmap: everyone wants it to work, but no one wants to work on it. Here’s an example: a senior executive at a leading SaaS company tells me that twenty people from different groups across the company show up for the “billing meeting”, where it’s decided how billing will integrate with core features. But no one wants to work on the billing team, creating those integrations.

Startups have capitalized on that by creating high performance, scalable integrations, solving hard technical problems like how to sync without putting excessive load on the underlying system.

Entire companies, such as our portfolio company Okta, for single sign-on, or Segment, for analytics, are now built on integration alone.

Integration companies, while not glamorous, can build market power by positioning themselves at the center of an ecosystem and creating an “ecosystem network effect”, whereby they become a de facto standard. Okta and Segment are both on their way to achieving this.

But most new applications use integration to gather, organize, and analyze data. They win the hearts of their users through great design. That’s no small challenge, given growing data sets, shrinking screen sizes, and ever shorter attention spans, which is why the concept of design has become a huge differentiator.

It works because it’s a win-win. Startups creating systems of engagement get users and revenue, by leveraging data in the systems of record. They also increase the data’s value, by using it more and adding to it. That makes the big software vendors happy, as (they believe) it increases their customer lock-in and helps them become more of a platform.

What’s not clear is whether this will continue. Large companies like Salesforce want to innovate through technology. For example, the center-piece at this month’s Dreamforce, its annual conference, is a new artificial intelligence (AI) initiative marketed as “Einstein”, which layers predictive models over its existing applications.

Conversely, once a startup’s product is being used every day like Slack, it may start keeping more information within it and over time wean people off whatever they were using before (Outlook, Sharepoint, etc).

The game-changer could well be artificial intelligence: if AI software could extract signal from the unstructured product feedback in Intercom or the sales forecasting information in Clari, the data in those systems could become more valuable than the limited fields captured in today’s systems of record.

But that’s a long way off. For current startups, the message is clear. Don’t try to be Salesforce to Seibel, Workday to Peoplesoft or Coupa to Ariba. Those battles are over, and won’t be repeated. Instead, use technology — integration, design, perhaps machine learning or AI — as your wedge into the market.

Play nice with existing systems, and then analyze how people are using your product. Feed that back into new product development and drive more engagement, ideally creating a virtuous cycle between usage and design that keeps you ahead of competitors.

For examples of who does this well, look no further than the large consumer companies. It’s no coincidence that the two most awe-inspiring enterprise businesses today (AWS and Google Apps) both have a consumer heritage.

That’s the winning strategy for today, and most likely tomorrow.

Featured Image: Ismagilov/Shutterstock