Few will argue these days that software has eaten the world. Software tools businesses led to SaaS solutions which are giving way to tech-enabled everything. What this means is that every industry you want to build software for is somewhere along this curve and every industry has incumbents.
Enter: The System of Record
When approaching a new market, new target customer, you will discover one or more of these incumbents has taken the mantle of "System of Record." For recruiters, it's the ATS, for Finance and Retail, it's the ERP, for Sales/Marketing, it's the CRM, for Support, it's the ticketing system, for Manufacturing, it's time and attendance, for HR, it's the HRIS, and on and on...
There is a religious fear and awe of these things, whether the software seems good to us techies or not. Why?
Well, for one, the customer is used to it. Incumbent software may suck, but it's sucked for a loooong time. Customers are comfortable with the way in which it sucks and have already built business practices around mitigating how much it sucks. You're also going to suck in some way, so why change?
But it also just happens.
I tell PMs they need to be architecturally technical and this is one of the lessons: moving a bunch of data around is hard. Heavily used packaged software is a data magnet and, eventually, that data becomes really important. That data is the business in a very material way and the idea of migrating it to another system conjures fear of loss.
Startups, of course, hate the System of Record. It's everything we're not, old and slow with terrible UX. Their Sales people are slimy and rude, ours are obviously clean-cut and thoughtful, and their Support staff would sooner spit on you than fix your issue.
So we often struggle to deal with the System of Record, even if that relationship is what eventually makes or breaks our business. In this post, I'll break down how to work with these large incumbents and their mind states around each of these decisions (as well as their customers').
1. Learning Coopetition
Startups are competitive by nature, they have to be. Every sale is a challenger sale because no one's ever thought of doing anything differently. Every YC batch may be loaded with upstarts coming after your market.
Big companies don't think like this. They want long term advantage and market position over their competitors, but know that they're entering polyglot environments. While startups are trying to land, the incumbent is 5-10 years into expanding. Vendors are always rising and falling in a company's tech stack and incumbents know partnering with competitors is important to maintain or grow their position.
So the first step is basically to swallow your pride. You may disagree with a Big Co. vendor or even see them as directly competitive, but joint customers don't care. You need to instead think about the end result for the customer, what that solution looks like, and whether that solution has value for more customers and your company as a result.
2. Partnering With Big Co.
Just knowing that you need to partner with a System of Record isn't always enough. Different incumbents are on different stages of their journeys when it comes to partnering. Workday's partner landing page used to effectively say, "No thanks!" Microsoft, in contrast, financially supports startups that work with their technology.
As a startup the success of the company may hinge on your ability to deliver a partnership whether the incumbent has a well-defined partner program or not. You need to understand their motivations and find a path to work together anyway.
Part of working with Big Co. is understanding how they see you:
The Shiny Toy - you'll initially be perceived as a Shiny Toy by any incumbent. Part of serving a niche is that the need you address will only be a subset of their customer base. Incumbents will have either disqualified this niche or believe it to be satisfactorily served by some half-baked functionality they have. The problem of being a toy is that the incumbent doesn't take you seriously. They aren't going to invest resources in even learning what you do. This can work with very open incumbents, like Microsoft or Salesforce, but others have gatekeepers. These gatekeepers have no incentive to say yes to every small fry off the street. Your best bet for getting around a gatekeeper is through their customers. A crescendo of support for an integration will be communicated through their field teams. Better yet, one influential customer might have ties directly to the BD organization. You may stay a shiny toy to the incumbent, but an API key and a marquee customer are a good start.
The Lynchpin - the most advantageous and dangerous position. The incumbent needs you to fend off a competitive attack from another vendor. These relationships start in the field, with individual reps wanting to position your capabilities along with their software. The incumbent may become a reseller or even OEM partner, with your tech embedded in theirs at some point. The risk of being a Lynchpin is that company knows they need you and might not want to pay what it costs to acquire you. So they might find a cheaper alternative or build directly competitive functionality. As a Lynchpin, your goal is to avoid becoming dependent on the single vendor. With three large incumbent partners, you're less vulnerable to the whims of any one. If the incumbent wants to acquire you, you can bring other parties to the table to deliver the best outcome.
The Enemy - the people running the System of Record have different operating realities (mission, values, and near term priorities) than you. Even if you look at their software and yours and see no similarity, they might. While working at an HR software company, I once heard from a consumer device company, "Sorry, your company's been flagged in our CRM. We can't work with you." Remember coopetition, though. If the incumbent has an app store, you'll frequently find apps by direct competitors. Though difficult, it is still possible to partner with incumbents who consider you The Enemy. Companies who want to attract development on their platforms, in particular, can't be seen as monopolistic. Customers are again your best leverage point. Here it's important to spell out your unique value and let the customer state why they want the two systems to work together. As someone currently paying for (and theoretically seeing value in) both, they might see a path to you adding value that's less frightening to the incumbent.
There are a lot of variants of these relationships, but I think these are the top three archetypes: they don't care about you, they love you (maybe in an unhealthy way), or they hate you. Thankfully, there is a path to successful partnership from each.
There are Business Development folks who can help you navigate these relationships. In my experience, BD folks are as valuable as their rolodex and track record of signed deals. There are many who have neither.BD folks will sometimes work on a contractual basis, which might be the best way to step into it.
3. Understanding the Solution Set
So you've gotten the green light, you get to start building something. You'll probably see a lot of opportunity to make the System of Record better while you use that data to enhance your own offering. But when you suggest process improvements, the customer freaks out, what gives?
What Customers Want
We talked above about the significant investment of data customers put into their packaged software. Customers also spend a lot of time training staff and building internal processes around how those systems are used. In enterprise, they may have made a conscious decision to elevate one System of Record to "Source of Truth." Committees surround these Sources of Truth to prevent changes being made.
They may like you a lot, but your short operating history gives customers trust issues. They have Word docs describing processes older than your company. They don't want you messing with the data and workflows of systems they've used for years.
What the Platform Wants
Different Systems of Record expect partners to integrate in different ways. Salesforce partnership folks want "native" apps, where the application's database resides in Salesforce. These platforms may promote apps that integrate through the API, but the highest tiers of partnership are reserved for partners who support their architecture.
If an incumbent vendor pushes their architectural message hard enough, implementing a different way might even be a barrier for your sales team. Employees at prospective customers will be seeded with questions to make your integration seem illegitimate.
This is not to say these platforms are perfect or that this is the best approach for your company, but understand that violating these norms is a hard road and better be worth it.
What You're Good At
That last part comes back to you. Just because the incumbent vendor wants you to build 40 different versions of one integration to their versioned SOAP interface, doesn't mean it's smart or in your best interest.
Your value prop might be improving data quality in Systems of Record. Getting the customer to be okay with you making changes to their data might be the whole ball game. Or you might just not have folks who get working with big box software. If you're in big enterprise, something like a Mulesoft or Jitterbit integration might be enough.
Just understand where customers and potential partners are coming from and be ready.
4. Sticking the Landing
So many partnerships are made or broken at launch. A partnership is treated like a quick experiment, maybe announced, weakly supported.
A partnership should be thought of like a brand new product. As such, it needs support from the whole organization, not just the BD team. An ideal partnership has a roadmap, specs, development support, and a go to market plan with targets and accountability.
It might feel like a throwaway integration, but that just means you need scaled down versions of your launch plan not zero launch plan.
The worst outcome for the company is a big splashy partnership that doesn't do anything. The dreaded Barney Relationship. Where customers don't see a ton of value and no one benefits financially.
To succeed in a world of large incumbents who've claimed the mantle of System of Record, you need to:
Learn Coopetition - working with partners who also compete
Partner With Big Co. - understand how to succeed based on how they see you
Understand the Solution Set - build for the platform, but don't sell your soul
Stick the Landing - launch partnerships with the same effort as a new product