Build vs Buy – how core is the feature?
This is the second post in our Build vs Buy series where we discuss the Learnosity team’s approach to the age old build vs buy dilemma; whether to build your application from scratch in-house or whether to licence the technology from a specialist vendor. There were varying elements in each scenario that contributed to the overall decision but when we looked at three of our most recent build vs buy decisions, two overarching themes emerged:
- How core the feature is to us.
Our first post covered all the decisions associated with cost, including often forgotten ones such as maintenance and support costs as well as sunk costs.
This post focusses on the importance of the particular feature and ultimately how it contributes to our competitive advantage.
Contribution to Competitive Advantage
This is a key one for us. Every successful company has a “thing” – something that makes it unique in the world. We’re an education technology company and even more than that, we’re an assessment technology company. We specialize in assessment and everything related to assessment: the authoring of the questions and activities, the assessment delivery and the analysis of the student responses and trends, we don’t do anything else. Basically, we know what we’re good at and stick to it – it’s our thing! In the words of our CEO, “we go a mile deep and an inch wide”. So when making the build vs buy decision something that influences the decision hugely is whether or not the feature or functionality is core to what we do.
To take a specific example we recently outsourced some of our monitoring to DataDog – a cloud-based monitoring service. At the time of making this decision, we already had a monitoring service which we’d built in-house. In fact, we’d spent considerable time and resources over the last 4 or 5 years building this awesome 24-hour monitoring system, which would send SMS alerts if servers went down or systems were reaching critical load. It was working well and we were pretty proud of the system we’d built to be honest. However, we came across Datadog and we realised that they were always going to do it better than us. Not because our DevOps team couldn’t build it – we are fortunate to have an amazingly talented team but simply because monitoring is “their thing”. Much like assessment is our thing, monitoring is theirs. It’s what they do and it’s the only thing that they do so they have a laser focus on this one small niche area.
Yes, we had to scrap what we had already built, and this was by no means an easy decision. We’d spent years building out this great system. It was doing its job so did we really need to replace it? Taking a step back and looking objectively at it though we realised that leveraging Datadog would do several things for us:
- Remove the need for maintenance & developer resources going forward
- Free up our DevOps team to focus on more mission-critical work
- Provide the opportunity to leverage other awesome monitoring services, that we would never have built in-house because while monitoring is important to us, it’s not our speciality so those features were always going to be on the “nice to have” backlog.
How customizable is it and could a vendor do it better?
This is another key one for us and it’s related to how core the feature is to us. In theory, we could build almost everything related to IT and software applications in-house, we have the IT talent to do so. But ultimately why would we? Why reinvent the wheel?
When making this decision we need to look at several things, the first being how integral this feature is to our offering.
Granted if we built it in-house, we would have full control over the evolution of the feature and could easily adapt it over time, adding features as requested by our team. There are a few downsides to this approach though. Having full control is generally perceived as a good thing, but there are cons. One of the big dangers of building from scratch is building what we like to call Homer Simpson’s car – car that does everything for everybody but excels at nothing. Yes, we can build the application to do X but should we?
Tying into this sometimes is the fear that an external provider may not provide us with the flexibility to build as robust or as customizable a solution as we need. Our specific needs, while not necessarily 100% unique all the time can still be very different to others in our sector. So a key thing for us when we’re looking at potentially outsourcing is how flexible and customizable the software we are purchasing is? Can we easily modify it enough to suit our unique needs or are our needs really so unique that the only possible solution is to build in-house?
One of our most recent strategic moves illustrates both how cost and, how core the feature is to us played a role in the build vs buy debate. We decided to leverage Desmos, an online graphing calculator, in some of our Math and Graphing products. We’d built much of this functionality in-house ourselves already so there was definitely a sunk cost involved, however we recognised that there were a few reasons why we should buy rather than build.
Firstly, graphing is their “thing”. We could potentially replicate the functionality but by the time we get to parity, investing a bunch of time and resources to do so, they’ll have progressed their offering even further. We’ll always be playing catch up so why reinvent the wheel?
Secondly, leveraging their expertise frees us up to focus on other more important things. Graphing and scientific calculators are absolutely an important component of the Learnosity offering, but in the scheme of things they’re one small tiny, albeit incredibly complex, part. Leveraging Desmos means we can dedicate our time and resources to other more core features of our offering – ones which will add more value to our clients.
These posts detail our first-hand experience of the things we consider when making a build vs buy decision. We hope you find it useful and we’d love to hear your thoughts and feedback. What do you consider when making the build vs buy decision?