How Much Integration is Enough Integration?

Why do businesses integrate software systems? There are many underlying reasons, but it always boils down to one goal: maximizing transparency and data usability across departments while minimizing the time investment and cost of managing data.

With this goal in mind, it makes sense that a business would want to integrate as many systems as possible – but is that always the best solution?

Native integrations, like those that come pre-baked into popular software systems (e.g. Salesforce, HubSpot, etc.), make pairing with another major system simple and clean. But, when you consider all the other systems the company may be utilizing, you run out of pre-baked solutions — fast.  It’s not uncommon for a company’s core toolset to include multiple solutions, all housing their own subsets of data. Common solutions include CRM, marketing automation, ERP, HR management tools, MRPs, help desks, email, website analytics tools, webinar platforms, data enrichment tools, and several more.

At scale, integrating a few of these tools together may be doable, but when data is rendered in so many variations, it becomes nearly impossible to build a single source of truth that gives context to it all.

Think of it like this: if you had three tools, you would need three pairwise integrations to sync data appropriately. However, as you add solutions to your toolset, the number of pairwise integrations increases exponentially.

Using the formula (n*(n-1))/2, where n is the number of tools you have, you’ll quickly find the number of integrations skyrockets the more systems you have.  We use this equation because when we have n systems, each having n -1 connections (each is connected to every system except itself), we get n*(n-1). We divide that by two because connection (x,y) and (y,x) is the same (for all connections), and we end up with n*(n-1)/2.

Remember that seemingly 1-to-1 relationship between three systems and their three pairwise integrations? Well, plug in this formula for 10 systems, and you now need 45 pairwise integrations to achieve the same level of transparency — whoa.

Not every system can integrate the amount of data your major systems can. You could try to partially integrate all systems, but the result will be a fragmented picture of your data. This makes the goal of transparency and lowered investment costs obsolete.

So, where’s the middle ground? Here are a couple of tips for your integration strategy.

1. Map and consolidate your data sources. First, define all the sources feeding data into your business. Then, analyze the importance of each of those systems. Many businesses can find at least one tool that is underused or has overlapping features with another system. The more tools, the more complicated the integration strategy, so consolidate where you can. One option would be to utilize third party add-ons with your existing systems instead. For example, you may choose to use a service management add-on with your CRM instead of purchasing a customer service portal. Or, eliminate overlap by consolidating to one tool. For example, if you’re using HubSpot for marketing automation and a separate chat service for your website, consider getting rid of the separate chat and using HubSpot’s chat feature instead. That eliminates one more integration and, since your data is already in HubSpot, it’s simpler to track.
   
2. Recognize that your integration strategy is highly dependent on your company’s maturity. If you’re scaling up, you’ll be adding more tools. If you’re a later stage company, you’ll have more teams, tools, and a bigger budget to consider.  If you’re a small business with a limited budget, you’ll have to strategize which systems make the most sense to connect. The current status of your company, and the company’s growth plans going forward, impact how you tie your data together.
   
3. Pick one system as your system of record for a particular type of data. Instead of connecting everything to everything else, centralize the data with one system of record. Let’s use Contacts as an example of a particular type of data. For most companies, your system of record for Contacts will be your Customer Relationship Management (CRM) system. CRMs typically have the most robust integration capabilities for managing contacts, which makes them a logical choice as a system of record. The data about Contacts inside this system should be FACT. This is where everybody goes if they want to know more about the contact. Each CRM takes extra steps to make sure data is uniform, consistent, and factual. Advanced CRMs deploy data entry strategies (ex: force users enter dates as “October 12” instead of “10/12” or “Oct. 12” or “10.12.19”) to get rid of duplicates or outdated information and establish it as the first entry point for any new customer information.
   
4. Run what you can in real-time. There’s a delicate balance to this. While you want your data to be as current as possible, you also don’t want to slow down all your systems by continually syncing every snippet of information. Be selective and choose only your most crucial data sets to sync in real-time. Which fields/functions are most important to the majority of users?
   
5. Choose an integration tool as dynamic as your data. – The Cloud is not the be-all-end-all for every business. Some companies use certain applications in the Cloud and keep others on-premise for regulatory reasons. Make sure your integration tool can handle on-premise, Cloud, and hybrid integrations so all your bases are covered.

If you’re not sure where to start with your integration strategy, let our team help you. We’ll review your needs and talk you through the pros and cons of your approach so you can build the best integration framework for your business.