Not all matching algorithms are made equal. Almost every other marketing or sales automation company claims to be performing lead-to-account matching algorithms, but are they enough to help you achieve your business goals?
The algorithm needs to be looked upon from different perspectives, and depending on your specific need the weightage of these may change.
Few companies do the email/website domain match only, while some other companies offer Company Name Matching Algorithm by removing legal suffixes and common words such as Inc, Corp, etc. A comprehensive matching algorithm often depends on a combination of rules and dictionaries. For example, LeadAngel’s matching algorithm has over 160 rules and millions of dictionary entries to look up. Rules should often consider following while matching:
All these rules work in isolation as well as in combination, yielding the best possible match. Though it is paramount to never miss a possible match, it is also extremely important to keep out the false positives. We typically recommend less than a 1% margin of error for both of these metrics.
The software must provide for an ability to customize the rules, the score, and the provision to re-run the match if needed.
Tie breaker is often ignored (but should not be) during the initial matching discussion. Lead-to-account matching is not an exact science. It gets even more complicated when there are duplicates and multiple close matches. Thus, the role of Tie-breakers becomes important. Oftentimes, it is more productive to match the leads against the most active accounts, where there may be a business case to match against the oldest or newest account. Sometimes, the most active account is selected based on the number of leads, contacts of opportunity tied to it. It is paramount to select the lead-to-account matching software that provides this feature, otherwise, it will come back to haunt one day. Also, look for a default tiebreaker if everything else being equal, it is generally acceptable to match a lead against the geographically closest account for better territory profiling and routing.
An enterprise CRM implementation involves heterogeneous accounts. The Account may be categorized by Parter vs Prospects vs Customer etc. Leads are usually matched against either partner accounts, or against the inside account lists. Matching leads to the correct account list helps in reporting, analytics as well as lead routing. Sometimes, leads need to be matched against a Virtual Account list as well. Thus, while selecting a software vendor for L2A matching, be sure to check for the capability to match against various account lists.
Match direction is rarely discussed until it is too late to discuss. Most of the lead-to-account matching algorithms are just that, they match lead to an account whenever a lead is created or updated. Think of a scenario where a lead is created, but found no matching account. Later on, after a few days, a matching Account is created. If the L2A matching software does not automatically do the “reverse-match” to link the account to lead, the sales team will end up missing critical information such as “other matching leads” from the account, or any of the accounts histories on the lead for that matter. The biggest drawback is that not all the leads that could be matched to an account are matched at any given time. It will result in incorrect routing, reporting, analytics, segmentation, and tie-breakers. It is advised that you invest in complete software that does “Lead to Account” matching as well as “Account to Lead” matching.
The matching should happen at the speed of business expectations, often governed by the lead routing speed. If your business expects the leads to be routed within 10 min, then a real-time trigger-based match would be overkill and expensive. Batching the lead to account match reduces the API usages. If the business’ lead routing is rather real-time (say ISR must call within 30 seconds), then the API-based match is desirable.