The adoption problem no one talks about
CRM automation has a dirty secret: a significant proportion of the workflows built in enterprise CRM implementations are either unused, producing incorrect outputs, or have been manually overridden into irrelevance within 12 months of launch. The automation is technically running — it's just not doing anything useful. The problem is almost never technical. The workflows work as built. The problem is that they were built for a version of the sales process that either doesn't reflect reality or that the sales team never agreed to. Automating a flawed process just produces flawed outputs faster.
Rule 1: Automate the process people actually follow, not the one in the playbook
In every CRM implementation, there's a gap between the official sales process (as documented in the playbook) and the actual sales process (as executed by the top performers). The official process has 12 stages, three approval gates, and a mandatory executive review at qualification. The actual process has 5 stages, with two approval gates that are rubber-stamped anyway, and no one enforces the executive review. Before building a single automation, shadow your top three salespeople through 5–10 deals each. Document what they actually do. Build automation to support that process — not to enforce the idealised one. Automation that enforces a process nobody follows just creates friction and gets disabled.
Rule 2: Automate the admin, not the selling
The automations with the highest adoption rates in every implementation have one thing in common: they save salespeople time without asking them to change how they sell. Automatic activity logging from email and calendar. Deal stage updates triggered by document opens or link clicks. Follow-up task creation when a deal goes stale. Automatic data enrichment from a third-party source when a new contact is created. The automations with the lowest adoption rates are the ones that try to change sales behaviour: mandatory fields that must be filled before a deal can advance, approval workflows that slow down deal progression, automated scoring models that override the rep's judgment on deal priority. These create resentment. Build automation to remove work from salespeople, not to police them.
Rule 3: Measure and show the time saved
Sales teams are sceptical of CRM investments because they've been burned before — time-consuming data entry, tools that don't work on mobile, processes that slow deals down. Changing that perception requires evidence, not advocacy. Three months after each automation launch, measure the actual time saved. Pull CRM activity data: how many tasks were auto-created vs. manually created? How many records were enriched automatically? How many stage changes happened without manual update? Translate these into hours saved per rep per week and present it back to the team. When a rep sees 'our automations saved your team 4.5 hours each last week,' adoption follows.
Found this useful?
We write about automation, software strategy, and engineering once a month. No spam.