Build vs. buy: Is that still a question in retail IT?

Photo: @kirsty via Twenty20
Aug 14, 2020

Through a special arrangement, what follows is a summary of an article from Retail Paradox, RSR Research’s weekly analysis on emerging issues facing retailers, presented here for discussion.

A lot of old-school thinking still pervades retail management when it comes to information technology as the “build versus buy” argument continues inside many retail headquarters.

On the “buy” argument side, most retail shops are working with application portfolios that are rigid, out-of-date and expensive to maintain. IT backlogs continue to grow and companies struggle to find budgets to devote more spending to value-adding capabilities rather than maintenance.

When it comes to implementing new applications, pushing past a “not invented here” mindset is critical to taking advantage of the latest state-of-the-art solutions.

Many (if not most) core retail business processes are found in every company. Retailers get no credit for managing them well, but they do lose points for managing them poorly. Decision makers should ask themselves, “Which business processes create truly differentiating value that consumers will benefit from and give us credit for?”

The good news is that today’s commercial solutions generally work; unlike the last generation (which were really frameworks more than complete solutions), modern applications are functionally rich and agile enough to be able to conform to retailers’ internal business processes.

But while commercial (and increasingly cloud-based) solutions can accelerate modernization of a company’s IT capabilities, don’t be fooled by promises of an “IT-less” future.

Some commercial solutions are built to easily integrate with legacy portfolios, but companies must address business process re-engineering needs. Eliminating process bottlenecks, reducing cycle times, automating exception handling and other streamlining goals should be a focus for IT teams when choosing a solution.

Proprietary solutions still play an important role. Retailers need to commit to use precious IT development resources for things that help the company deliver truly differentiated value to consumers.

Finally, it’s important to remember that Rome wasn’t built in a day. Many retailers have a significant backlog of work to do to bring their IT portfolios and the processes they support up to date. They should expect to live in “dual mode” for years to come.

DISCUSSION QUESTIONS: Where do you land on the debate over whether retailers should develop their own software applications or use commercially available solutions? Should increased accessibility of commercial solutions recast the role of IT teams?

Please practice The RetailWire Golden Rule when submitting your comments.
"You want flexible, modular, open architecture that plays well with other applications – not something locked in a proprietary data format with highly expensive exit costs "

Join the Discussion!

19 Comments on "Build vs. buy: Is that still a question in retail IT?"

Sort by:   newest | oldest | most voted
Suresh Chaganti

As someone who led IT organizations in retail, I have first-hand experience of this question. The reality is, innovation comes at a much faster pace than what your internal IT team can keep up with. But off the shelf software isn’t great at keeping up with innovation either. Any software, built or bought, becomes “legacy” the day after deployment.

So it is always a trade off between build and buy, short term and long term. Sometimes decisions are based on technology, other times resource availability, functionality, or financial concerns.

What you want is a flexible, modular, open architecture that plays well with other applications – not something locked in a proprietary data format with highly expensive exit costs.

Dave Wendland

To be honest, I think retailers should focus on their mission of delivering value to their shoppers at every touch point. And frankly, it is hard to do that if they are also trying to maintain mastery at IT. My recommendation — contrary to what I’m sure will be other views expressed here — is to find a packaged solution for most applications and then apply customization layers that meld the application with unique operations.

Speed to market has never been more critical. Retailers MUST focus on what they know and do best leaving things like software development, the science of assortment planning and category management, digital asset and content management, supply chain traceability, and other functions to seasoned experts that can bring immediate value and fill the voids.

Richard Hernandez
Richard Hernandez
Director, Main Street Markets
1 year 11 months ago

Commitment to IT infrastructure is essential – it always has been.
The problem is that no more applications developers are being added to really develop or enhance in-house systems to take them to the next level. You work with what you have which is why the buy option remains a strong choice. No application runs right out of the box as we want it to – data has to be reviewed and cleaned, processes and rules have to be set in place and a set of objectives (short, medium, long) have to be set in place to measure the success or lack of abilities of the application. There should always be a priority list of what can get the most bang for the buck (pricing strategy, vendor portals, ad systems) and another list of things that will take longer to prove out (warehousing, cloud-based accounting, etc).

Bethany Allee

It’s not a black or white answer. The perspective is important, because there are many areas where proprietary IT solutions have been perfected – and in these instances, retailers should use that technology. For example, security. I would advise retailers to leverage outside security tools for a multitude of reasons.

There are several areas where retailers have specializations or proprietary elements in their operations, elements where IT solutions can’t accurately capture, recreate, or service. In these instances, it likely makes sense for retailers to build out their own solutions — and this is where services come into play. IT services is one of the fastest growing segments within the technology space. Service providers can help retailers build solutions faster and better, without the long-term expense of an in-house development team, network, application…

Given the volatility in today’s world, it’s critical that retailers minimize expense. Instead of focusing on proprietary/non-proprietary applications, retailers have an opportunity to step back and leverage proven IT services to gain proprietary and non-proprietary excellence in all aspects of their business.

Bob Amster

The pros and cons can be debated at length because there are multiple factors affecting a decision. My bottom line, and it has been so for decades, is to buy, avoid customizing, and stay no more than one release behind whatever is being offered. Period, the end. Only if a business has INVENTED a heretofore non-existent secret sauce will there be no commercially-available solution to support it. In that way, one develops one’s own until more commercially-available solutions flourish.

Ryan Grogman

Buy for base functionality, configure for unique business rules, and custom develop proprietary differentiators. It’s not always as easy as that, but it’s a good guide when assessing the buy vs. build decision.

Ken Morris

This is like breathing — best-of-breed off-the-shelf software or roll-your own-custom development? The only answer is that it depends. I’ve had to dismantle too many out of control development projects that took multiple years to develop, at a ridiculous cost, with yesterday’s specifications, that were behind the packaged software that was available on the market for a fraction of the price. My rule of thumb is that if it provides a strategic advantage and you can’t find an open architected solution to buy then invest in the development.

Gene Detroyer

After Suresh’s and Dave’s comments there isn’t much for me to add. This morning I was talking to an IT guy on this very subject, but how it applied to hedge funds rather than retail. He noted, “Once upon a time IT was about programmers, today it is about data scientists.” In essence, if a retailer is not big enough to include a data scientist on their IT team, go outside. The depth and scope and speed of IT development today makes it extremely hard for a non-tech company to progress with the tools that are available. If you are large enough and DEDICATED enough, build your IT department. But the cost of the talent may make it the most expensive talent in the company.

Zel Bianco

Retailers have enough issues on their plates to worry about. Technology changes so fast and there is just no way for IT to keep up. Retailers would be better served working with their solution provider to adjust and customize as needed. We have been developing category management analysis and presentation software for over 20 years and always need to update, pivot and respond to customer needs. I have many examples of customers who tried to build internally. They spent a good deal of time and money to come to the conclusion that they could have met their needs much more effectively, in a faster time frame and at a much lower cost had they bought. Those companies are world class CPG companies who became some of our most loyal clients!

I will apply the pareto principle (80/20) rule. The advantage of leveraging third-party solution providers is scale, innovation and their R&D. For example do you really want to develop and AI (ML/Predictive Analytics) platform with your own data scientist? No doubt you can move faster with third parties who are focused every day on their technology and as a retailer take advantage of their platforms — especially if they are truly cloud native. I always ask retailers, what is your secret sauce and what is it that you do that makes you different and that no one else can do? That is where you focus your internal IT teams and business analyst (the 20 percent) and let the third-party software and solution provider focus on the 80 percent. I think that so many retailers are sitting on monolithic and bifurcated applications to run their business and they are simply stuck. Hence the need to move to platforms that are cloud native with a common data model built with integration at their core. Salesforce buying Mulesoft… Read more »
Sterling Hawkins

I don’t think the question is one or the other. IT organizations cover so much territory that there’s ultimately probably a need for both depending on costs, timing, existing tools, roadmap, etc. The more modular and flexible the better. Technology is always a means not a purpose. Keeping the customer front and center will generally lead to better decisions around not just technology, but everything in the long run.

Shep Hyken

It’s an easy choice for me. I’m either a retailer or a software developer. If I’m a retailer, I should be able to find existing software that offers the flexibility of customization and growth. I can pay my own developers to create and maintain or I can pay companies who specialize and devote 100 percent of their effort to software that will work for me. My choice is the latter. That way I can focus on retail.

Craig Silverman
1 year 11 months ago

In today’s world, no retail organization can fully build software internally. For a retailer, it is very difficult to keep pace not only with the changes on the tech side with cloud and AI/ML but also with core transformation in their business models with omnichannel and its impact across the organization’s process and IT. What retail organizations need to seek are not just software solutions but true partnerships that integrate and evolve as they do. Many forward thinking organizations have shown outside partnerships are extremely strategic to their business, so much so that there’s a new trend in acquiring those capabilities outright e.g. Nike’s acquisition of Celect or McDonald’s acquisition of Dynamic Yield.

Perry Kramer

The answer for larger retailers is both. For smaller retailers it is more than likely buy. The IT focus for buying should be on purchasing modern software that allows the retailer to extend the application with in-house developed solutions (or contracted resources) with the features that provide speed to market for new functions and enables an experience that is distinguishing to the retailer. To do this they first need to settle on an overall technology architecture that is, agile, extensible, and services-based. In parallel retailers need to take this approach to replace the legacy systems that many are spending up to 70 percent of their time supporting with little additional value to the business. Freeing up these resources will allow the retailer to invest in extending the core products that are readily available in the software market.

Andrew Blatherwick
It is true that many retailers are still struggling with out-of-date technology that is not helping them maintain best practices or gain competitive edge. If they do not invest, they remain locked into old technology that is not only costly to maintain but very inefficient. One of the main problems is retailers would rather adapt industry leading software to fit their outdated processes than change their process to maximize the benefits of any new technology purchased. The Build vs Buy argument is relevant because often they would build to fit their processes and not move on. Rarely do they look at the whole process change and update before they start to build. The argument you build if you have truly competitive edge ideas that will take you ahead of the competition is very valid, but reinventing the wheel is not a value adding use of resources. Some IT professionals do not help their retail companies because they are more interested in their own kingdom than delivering what is necessary for the operations team. A recent… Read more »
Gib Bassett
Seems to me that whether it’s building tech or building a storefront, you create unique things when it must be core to your value proposition – something your stakeholders highly value and associate only with you. You instead buy things that tend to be common among your peers, and tune these to your business needs around the edges of the core use case. The things you build uniquely, of course, must be architected to work smoothly with what you buy. You must have flexibility and interoperability. Easier said than than done, that’s the point. Every retailer comes at this problem with different baggage and considerations. It requires leadership and frankly a great analytics capability. I’ll bet there are few retailers who actually know factually what truly matters to their business such that they must build it themselves. Wouldn’t it be great to focus tech development investments on the edges of the core business processes that every retailer needs, tune them to meet the needs of your customers, and create new solutions that your customers appreciate,… Read more »
Peter Charness

It is still a question, and it’s one for which there is no good answer. The answer “buy and stay vanilla” sounds good, but let’s face it, some of the fundamental solutions a retailer can buy are old enough to vote, with some good looking lip stick. Newer SaaS based solutions can be less functional than their predecessor on premise solutions, and not as configurable.

I think Ryan’s answer is great advice. Implement the mainline of the solution as is/or configured as needed, and build add ons for edge cases and differentiation to leave the base solution “clean and neat.”

Ideally that base solution will have modern integration capabilities that supports this strategy. Good news is that there are newer low code/no code frameworks out there, or even emerging AI frameworks that can generate application code. These make the strategy of add – beside more and more feasible, even for smaller shops.

Ananda Chakravarty
Amazon $23B. Microsoft $19.3B. Oracle $6B, Salesforce $2.77B. Adobe $1.93B. IBM $5.99B. Just a sample of tech spend on R&D (source: respective companies and Statista). Even smaller players that can spend on developing new software like Shopify spends over $355MM on R&D. The tiny tech startup is also dangerous with new ideas and 100% of staff on their new technology. These options will outgun retailer IT teams almost every time (exceptions — Walmart, Target, et. al. — whose real advantage is not just tech) The determinant is space and software maturity. It doesn’t make sense to invest in designing your own software for POS or CRM as these systems are quite mature and well developed. Even the unique considerations of your retail business would most likely require tweaks or configuration changes rather than starting from scratch. Add to that the fact that these systems are now mostly available via SaaS with continual updates, you’ll quickly fall into a technical debt trap by pursuing your own solutions. Places where it might make sense are in IoT,… Read more »
Casey Craig

Yes, it is still a question. For systems of record (finance, HR) the answer is always buy. If you want to truly differentiate or be innovative then the answer is build. Look at the business capability you are supporting and if it is like all your competitors then there will be a COTS solution that can support it. If the business capability is a differentiator in the market or a completely new concept then building will most likely be required.

"You want flexible, modular, open architecture that plays well with other applications – not something locked in a proprietary data format with highly expensive exit costs "

Take Our Instant Poll

Are you a bigger advocate of retailers developing in-house developed software applications or using commercial solutions?

View Results

Loading ... Loading ...