Build vs. Buy: some thoughts, and the "Goldilocks" option

Read this if: You are on the technology/digital product side of a large enterprise, or you sell to someone who is.

• How the best companies that I work with think about "buy vs. build" for their product roadmap
• The "Goldilocks" option, and when to use it

Let's say I need something for my business. Do I build it myself? Or buy it from someone else? An age-old question, and I am going to show why the answer matters for you whether you or the buyer or the seller. I am also going to talk about the "Goldilocks" model that is becoming increasingly popular.

Pretend I am Dave Digital, Head of Digital at AcmeCorp, a large multinational company.

??.

Dave is the personification of all of the different conversations I've had over the past 18 months with leaders in the digital, product, and innovation space. If I recognize a business need, I have some options...

• I can build the solution myself. Or have a consultant build it for me. See that one a bunch, too.
• I can buy a solution from someone else.
• Or I can buy someone else who has the solution already. (There are various pros and cons to option C, but we aren't going to worry about that option in this article. Long story short: if you are Facebook and you can buy Instagram for $1B, do it.)

So what do I do? What are the pros and cons of each approach? And when is the right time for each?

How to decide?

I am fortunate, as I get to speak with digital, strategy and innovation leaders at some of the best companies in the world on a day-to-day basis.

And you know what they all have?

A framework for making this decision. Sometimes it's rooted in company culture. Sometimes it's a logical decision tree. But there is always something.

Whether you are the enterprise, or you are someone who sells to the enterprise, this is important. Here are some of the key takeaways:

• If it is mission-critical to my business, it needs to be bespoke to my business. Out of the box doesn't typically work for large enterprises, which is why we see players in the System Integrator space make the hay that they do. Out-of-the-box isn't a thing if you are vendor playing in the large enterprise space. This doesn't rule out the "buy" option, but it means I need a vendor who is set up to be flexible and accommodate my needs.
• I don't want a middleman between me and my customers in my primary engagement channels if I can help it.
• If it is core to my product, I want to own it. It puts me at a huge risk if my product is dependent on an outside vendor.
• If it is a commodity or a solved problem that is not core to my product or business, I will buy it.

Really at the end of the day (and in an over-simplified manner), Digital Dave thinks about it this way:

1. CAN we do it ourselves? If the answer is yes, proceed to the next question. If no, then we have to buy it.
2. SHOULD we do it ourselves? Is it worth it?

At this point, lots of variables are weighed out: is this critical or ancillary to our product? What is the cost in dollars for each path? What about in time? How many people will this tie up? And then you have branding issues, corporate philosophy issues, etc. Lots of inputs here. A hallmark of a really good company is that they are really good at figuring out if it's worth it to build themselves or not. But it's a hard question, to be sure.

Distilled all the way down: I'm going to BUY a solution if it's important and either...

A) ...I can't do it as well as a vendor...

EXAMPLE: Google is better at getting my product in front of people's eyeballs than I am. No matter how badly I wanted to do it myself and how many resources I am willing to throw at it, Google will still be better. ???♂?

OR

B) ...the investment to build it myself makes the juice not worth the squeeze.

EXAMPLE: Could I have a team build a CRM system and integrate into all of the other services I use on the sales team? Probably. Actually, definitely. But why do that, when I can buy Salesforce? Spinning my wheels to solve a solved problem is a waste of my/my team's time.

SaaS sellers out there - 9/10 times you are in bucket (B) if you are talking to a large enterprise. It's not a bad thing, just be aware, and position yourself accordingly.

*Side Note: if you are "Option A" as a vendor, congrats. But if you can find someone in your target industry who is historically a "we don't buy, we build" shop, and get THEM to buy your solution because they simply aren't able replicate it themselves...? Double congrats. It's a goldmine.

Clinc was fortunate enough to have that scenario play out with our very first customer, and we are still getting calls 3-4 years later from other companies in the space saying things like "reaching out because I saw company X used you guys and couldn't believe it! Love to set some time to see what it is you guys do!" $$$$.

A real-life example

Back to the story here, one important point to make: The "is it worth it" question is going to be answered differently based on the company and their business model. Real life example for you:

The Point of Sale (POS) system in a restaurant is critical, and everyone would agree with that statement. McDonald's owns their own. Burger King and Taco Bell don't, their franchisees have a few different options to pick from.

Pros of Owning: control and speed. In particular, McDonald's can decide they want to support 3rd Party Delivery without requiring the employees in the store to deal with tablets and such, so they can provide someone like UberEats a direct POS integration that they know will work across their whole system. That is incredibly powerful.

Cons of Owning: huge costs to get that set up, and to support it perpetually. It also might not be viable, based on the agreements that the brands have in place with their franchisees. AKA, the business model might not support it.

In this case, it's worth it for McDonald's. In the case of Burger King or Taco Bell, they either a) couldn't own it, because of their existing franchisee structure, or b) decided it wasn't worth it to own it, and allocated resources to other projects.

So the point is, there is some wiggle room here on what "worth it" actually means for your company or your prospect.

If you are reading this as a seller, this is your big takeaway: you need to understand what bucket you fall into:

• "A" (you need me because you can't do it without me)
• "B" (you need me because it's too much work to do it yourself and you have better things to be doing with your team's time).

The "Goldilocks" Option

If you recall, at the top I mentioned the "Goldilocks" option.

And that is the Platform approach. When I say "platform" here, I don't mean it in the ubiquitous way that that every SaaS vendor in the world uses it. I mean a tool that lets me build/ configure/ customize my own solutions, unique to my business and business needs.

This is the approach that a lot of players in the various subgroups of AI/ML have used to great success.

Why?

Really, it plays well for two reasons.

1) 3rd party vendors provide advanced technology that is oftentimes more cutting-edge than anything the company could create themselves.

2) AI tends to require a huge level of customization, not just in terms of the use-case or problem you are solving, but also in terms of the data you use to train your model.

Instead of buying a product, I buy a license to a platform, and either my team or one of the many professional services firms out there builds the product bespoke to my needs.

Let's put a bow on this. I (as Dave Digital from AcmeCorp) want to incorporate Conversational AI into my app to offer a convenient channel for customer support that keeps my call center clear. What can I do?

My (simplified) choices

A. Build it myself - which requires a lot of horsepower on ML side. And consider our two test questions:

• Can I do it myself? Some companies can, but most cases, definitely not. And usually not as well as some of the vendors in the space.
• Is it worth it do it myself? Again, occasionally. I might need domain specificity or some unique feature that isn't broadly available in the market. But far more times than not, the answer for most companies is "No, it's not worth it."

B. Buy it - this is an option that we see a lot of companies take that are maybe a tier below the largest players in a space. It's the fastest path to market, but definitionally you will not have a unique solution

C. "Goldilocks" it, Platform-style - this is by far and away the most frequent approach that we see the largest players with the most resources take. It allows them full control and customization, while avoiding reinventing the wheel, and it also allows them to leverage best-in-class tech that they might not otherwise have access to.

The considerations here if I am evaluating a platform are: ease of use, power of the engine driving the platform, and how truly customizable my "products" actually are.

Lots of options and approaches for digital teams to use out there. Even if you decide the platform path is right for you... you might not discover it's not the right fit until you've failed to bring something to production.

For instance, I spoke with a Conversational AI leader from a large financial services company the other day. He and his team had been going down the platform path with an external vendor.

But the vendor didn't supply the level of domain-specific customization he needed. They are now starting to build their own experiences in-house, because the external platform couldn't scratch the business itch. So do your homework in all three cases, but especially if you go with the platform approach. Not all platforms are created equal.

TL;DR:

1. Doing it yourself provides the most control, but eats up the most resources and you might not have the same capabilities as external vendors.
2. Buying a product is the fastest path to market, but is the least customizable, and is likely not a source of uniqueness.
3. Using a platform (the "Goldilocks" approach) is a best-and-worst-of-both-worlds solution. It provides you with an organizational capability, not just a solution for one specific problem via a product. It allows you to use the best tech available in a fully customizable manner. But it also requires significant resource investment from either a) your team to get trained on the platform (in the form of both dollars and their time) or b) a system integrator/consultant (in the form of dollars, often significant ones) to build/maintain the solutions and manage the platform.

??