How To Spot Ineffective Data Integration Solutions
Fool me once, shame on me, fool me twice…
Wouldn’t it be nice to not be fooled at all?
Data integration is still a fairly young industry. With any maturing space, it gets harder to sort the worthy solutions from the bandwagoners and lesser solutions. The more popular an industry becomes, the more confusing it gets for those on the buying side of the table. And with data integration as an industry projected to be worth $24.95 billion USD by 2026, we could just be seeing the tip of an exploding technology industry. That being said, it is imperative for merchants to pick the best data integration solution that meets their business’ needs as well as works effectively.
Coupled with the growth arc many companies go through, it’s not surprising that by the time merchants reach VL OMNI that they’ve already had a number of experiences with other data integration solutions or approaches. What persists is the number of mature merchants who are still going through a number of ineffective data integration solutions. Consequently, encumbering their growth and slowing their agility to respond to their own market changes.
While making mistakes is valuable to lessons learned, going through multiple applications in any vertical shows that there’s a bigger problem to solve.
What is it? Merchants are being sold on data integration approaches that they never should have been evaluating from the get-go because it was never going to work for them.
We’d all like to make less mistakes. In this article, we’re taking our 26+ years of data integration experience and sharing some valuable lessons learned on how to spot ineffective data integration solutions before you make costly investments and long-term commitments that turn into never-ending headaches.
We have identified four signs to help you spot ineffective data integration solutions:
- It was built on old tech
- It does everything! (AKA, the ‘Universal’ data integration platform)
- They pass work off to you
- Does the Future Matter To Them?
Read on to learn more!
Sign 1: It was built on old tech
Have you ever logged into a portal for an application and… cringed?
Data integration as an industry is evolving at an incredible pace. API-centric data integration has only been around for a decade or two (unlike EDI, which VL OMNI also does, which has been around forever by comparison’s sake). As such, new programming methods, interfaces, and best practices are still being developed and launched. The technology powering your data integration platform behind the scenes is important, but you, as a user, will likely never get a peek under the hood of what’s powering the integration platforms you’re evaluating.
The interface you as a user interact with, though, can be telling. Data integration, again, is a young industry, so there’s no excuse except a lack of innovation and iteration for a data integration platform that looks like a non-ironic retro throwback. If the data integration platform is built on cutting-edge best practice technology, their user interface should reflect this investment in the applications’ own technology. If something feels off, it’s a sign to ask some hard questions about what’s powering the integrations. And then get your pen and paper ready.
Most data integration providers worth their salt are pretty open about their technical capabilities, especially when asked (be skeptical of those who don’t want to share).
Signs that the data integration platform/partner is built on best practice and driven by new developments include:
- Ability to work with GraphQL, REST- and SOAP-based APIs. GraphQL is potentially the most cutting edge of the three, with REST being preferable to SOAP-based APIs.
- The ability to use webhooks, enabling the movement of data in real-time.
- True serverless technology, meaning the platform is powered by microservices and enabling the integrator to run integrations in parallel and not just in a linear manner. True serverless integration uses microservices, full stop, and not just hosting their services via a third party like Azure, AWS, or Google Cloud.
When it comes to indications that the integration platform is built on old technology, it gets a little more complicated. In general, there are some criteria you can take into consideration.
Some warning signs the integration platform is built on less-than-stellar technology include:
- Technology that was built around the turn of the century (2000) or earlier, like the Commerce One framework or ‘middleware’ (an old term for on-premise data integration technology). Old integration technology will reduce your business’ ability to be agile and ability to scale effectively to meet demands. You can expect issues like loss of data due to API throttling (maxing out servers and crashing them) and more.
- A linear approach to integration, meaning data transfer is dependent on other conditions. Workflows are not run in parallel, so one error along the way takes the entire connection down.
- No true process for error troubleshooting and reconciliation,other than maybe ‘we re-run the integration/connector’. Don’t be fooled — this is a manual process.
- And as a bonus, the user-facing interface for an integration platform is also worth exploring from a security standpoint. In the age of leaks, hacks, EU-GDPR and more, data security is of the utmost importance to you and your end users.
Sign 2: It does everything! (AKA, the ‘Universal’ integration platform)
While it might feel like we’re living in the future, we’re not quite there (yet).
What VL OMNI does is technical, complex, and involves lots of specialty knowledge and experience. Without getting too technical, APIs — application program interfaces — aren’t all the same. Different APIs are built differently and can do (or can’t do) different things. Some applications have very limited APIs that aren’t well documented (making it hard for integrators like VL OMNI to do what we need to do). On the other hand, some applications don’t have APIs at all and need different processes to extract data.
With all these variables, limitations, and conditions, how could a universal integration platform exist?
Simple: It doesn’t.
Sign 3: They pass work off to you
Does a self-serve portal make you feel like you’re in control and saving money?
If you said ‘yes’, here are two follow-up questions:
- Are you an expert in that technology?
- Do you work for free?
If you answered ‘yes’ to both of those questions too, then you can skip this section! But if you answered ‘no’, then keep reading.
A self-serve data integration portal may feel like you’re gaining (or regaining) control of your company’s automated data movements. However, these platforms are typically designed for agencies managing small customers, or for mid-sized merchants. In most cases, merchants who have grown out of lesser plug-and-play integrations but aren’t yet ready to make the jump to a managed solution like VL OMNI’s.
Self-serve portals are limited in their use because they have to be a) simple to use and b) serve the lowest common denominator merchant to gain the most ROI for the platform itself. They have to be simple to use, as the users aren’t programmers or developers, but rather laymen. Functionality will be reduced as a result, as will be the configuration of the integration to your business’ exact needs. These portals also need to have a decent market of users to support it, meaning if you’re looking for support or troubleshooting, you’ll likely be left waiting. Portals aren’t strong on customer support — they’re more like a factory churning out basic integrations.
The fact of the matter is that your time costs money. Outsourcing a complex need like data integration to experts will ultimately save you time and money, because integration experts like VL OMNI can do what we need to behind the scenes and configure all project needs much faster. Our managed service means we take care of the dirty work of setting up and any troubleshooting, and we’re customer service-forward. You get the VL OMNI Dashboard, but do not have to manage, field match, set business rules, or worry about changes — that’s all on us, leaving you to do what you do best: grow your ecommerce or multichannel business.
Sign 4: Does the Future Matter To Them?
We started this article off by starting with old tech, and this final section brings the conversation full-circle. One final way to spot an ineffective integration solution is looking at the platform in a future-forward way. What features does the platform have now, and what’s on the roadmap for the future? The answer can be hugely illuminating.
Analytics is a great litmus test to try on any and all integration platforms you’re evaluating. Most applications regardless of their vertical have some sort of analytics plugged in; which is a best practice. With data integration platforms, the presence or absence of analytics as a part of their platform exposes their technology strategy. Those without analytics or with basic analytics only aren’t using a centralized database. This likely means they’re using a linear approach to integration (Point A to Point B only, and only in one direction), which causes interdependencies that can take an entire connection down with a single error or issue.
In contrast, rich analytics with explorable depth indicates a more sophisticated integration approach. This includes holistic analytics, channel-specific analytics, and analytics not available via other portals, channels, or trading partners. Rich analytics are only possible with an API-centric forward approach that includes a centralized database.
These bells and whistles aren’t just ‘nice to haves’, either. They’re often indicators of the future development and potential of the integration platform, too. An inability to meet today’s standards in any web portal indicates issues today that are only going to compound tomorrow. Following our example on analytics, no analytics means a lack of data to power said analytics, AKA no centralized database and likely not a heavy reliance on APIs within the integration platform’s own technology. The implications for a platform like this are that they’re not API-centric and in no way ready for new developments like headless commerce or blockchain.
At the end of the day, the more experience you can collect on data integration, the more of an expert you’ll be on advocating for what your business really needs out of an integration solution. At VL OMNI, we’re up-front and honest about advising merchants who come to us. Even if it means pointing them towards a different integrator. We’re in it to build businesses and help merchants, and not just to turn a quick dollar.
We started this article off with a well-known proverb, and our honest goal with this article and others is to help the integration industry mature through quality thought leadership. But if I can leave you with anything on how to spot ineffective integration solutions from this article, it would be with two final proverbs:
“If it looks too good to be true, it probably is” and “Trust your gut” (and ask questions)