Software: to build or not to build? That is the question

This is a guest blogpost by Kevin Hurd, Founder & CEO, Digital Hive.

In 2015 Microsoft’s CEO Satya Nadella famously said: “every business will be a software business”. And it’s hard to argue that when companies like Uber, Airbnb, and JustEat are disrupting, then dominating, established, seemingly unassailable industries. If you’re under pressure to transform your company into a software business, you’re not alone. The big question then becomes: to build or not to build?

In every disrupted industry, one software trend – analytics – is separating the good from the great. Arming employees with fast access to reliable data for smart decision-making is essential for any company seeking to outflank competitors. It follows that many in-house development teams are evaluating whether to embed analytics in existing applications that people use every day, like intranets and CRM.

When Nadella made that prediction, he clearly didn’t want everyone to build their own software. Total control can be alluring, but there are many spanners that can land in the works: leadership changes, security issues, data privacy regulation, and constantly evolving technology versions, protocols, and standards – to name but a few. We have a whole developer team dedicated to staying current with analytics and BI software including ThoughtSpot, Cognos, Power BI, and the many others with which we integrate our intelligent portal.

However, I want to focus on one consideration that trumps all others – the user experience (UX). Unless you tempt users to adopt your software by offering an exceptional UX, all the beautiful coding in the world won’t matter. And these days, with millennials dominating the workforce, people won’t be comparing your software’s UX against those of SAP or Oracle; they’ll compare it to Google, Amazon and – the inspiration for our software’s UX – Netflix. Modern software is built for short attention spans, and expectations for instant gratification, personalisation, and device portability.

Adoption is top priority for the many companies working to provide analytics to business users. Gartner’s 2017 report on BI and analytics pervasiveness saw adoption rates “jump” to 32% after flatlining around 20% for a decade – but remains unacceptably low. The main obstacles we see to adoption are UX issues: users can’t find exactly what they’re looking for (navigability), fast enough (speed), without IT support (self-service). Today’s the bar is higher: users want to personalise their environments, get recommendations, and use search and voice commands.

To create UXs that drive analytics adoption, companies need specialist expertise. Even if your company has the funds to invest millions in R&D, finding and retaining ‘rock star’ quality UX designer talent is a bun fight. With a global skills shortage in UX design, LinkedIn has cited UX design as one of the top skills to learn in 2020 analysing the data from 660+ million professionals and 20+ million jobs.

As with most things the buy vs build decision all boils down to what your goals are. If the software you need is really specialised, not available on the market, and has few external dependencies, building makes sense. However, if your goal is widespread user adoption of a service like analytics that supports digital transformation, then buy – or a hybrid approach where you can do custom development with an SDK – are the ways forward.

A note of caution: what starts as a project to solve a specific problem can morph into something much more complex. Many of our customers started out needing to build a simple intranet to host company information. Then employees came to expect company information in the form of analytics, in (close to) real time. That means having to connect to data and reports from analytics and BI tools – a much more complex proposition to build and maintain.

So, the question shouldn’t be ‘can we build it?’ but rather ‘can we build – and maintain – software that people will use’?

Data Center
Data Management