How do you prepare for all the swirling ecosystem changes for the next couple years and beyond? 

Think of your CRM as having the ability to perform a specific sport, with all the aspects of that sport mastered. But now you must prepare to become good at multiple sports. 

Well, just like getting yourself ready for a triathlon, the first step will begin by measuring where you’re at today.  

Out with the Old
Before you can move forward, you need to clear the deck of technical debt that will slow you down. 

• Identify end of support and retirement items.
  It’s critical to list all the components that will not have support going forward or have already reached their end of support i.e. Salesforce Workflow Rules, Live Agent chat, etc. 

• Perform a comprehensive health check.
  Find the aging API versions, enforce your data archive strategy, update installed packages, etc. 

• Ensure security and privacy compliance.
  Has your metadata architecture kept up with GDPR, CCPA, and other regulations?
  Have you solidified the process for onboarding and offboarding users? 

Are your permissions configured on Profiles, or within Permission Sets in Salesforce? (If you said Profiles, you’ll want to revisit your future strategy). 

• Data fitness.
  Everyone bangs the drum of having good clean data, to use for future AI grounding. This KPI is only going to continue in importance, and you should already be tracking your current state.  Duplicated, orphaned, and incomplete records will erode trust in all your measurements. 

Two Change Drivers: Scale and AI 

Two forces are converging: business scale and AI capability. Manufacturing firms are consolidating acquisitions into a single CRM. Retailers are unifying online and in-store data. Universities are removing walls between department silos. At the same time, generative AI promises to automate qualification, summarization, and next-best-action recommendations, but only if your org can deliver clean, permissioned data in real time.​ 

Thriving orgs will be those that treat their CRM as a composable platform, not a monolithic application. That means modular architecture, event-driven integration, and strict data governance.  

The orgs that struggle will be the ones pretending it’s still 2018. Along with no test coverage, no change log, and no idea which fields feed downstream systems.  

Architect for Flexibility and Scale 

Future-proofing architecture is going to be a modular, composable approach. With domain areas such as Sales, Service, Finance, etc. These domains should communicate and share but not be tightly coupled and built in such a way that is brittle and easily broken.  

If a new acquisition needs to go live next quarter, can your architecture absorb it without a six-month rewrite? 

An intentional Developer Operations (DevOps) strategy and governance will also prevent integration sprawl and custom code when configuration is possible. If integrations and configurations are invisible to your DevOps pipeline, pieces will certainly break when you least expect it. 

A modern DevOps will have source control for all metadata, and work towards automating more, such as test coverage and even deployments.  

Speed of change	Slow; every change risks breaking unrelated features	Fast; isolated domains deploy independently
Merge/acquisition	Requires full rewrite or separate org	Plug in new domain via events and APIs
Testing	Entirely manual, limited; no standard	Automated assistance, modular, comprehensive
AI deployment	Not scalable; data permissions tangled	Controlled; strict access policies, ROI defined
Limits	Frequent collisions	Isolated or tracked by domain
Cost of ownership	High; every change is expensive	Lower; standardized patterns reduce rework

 

 

Use Case Examples 

Manufacturing	Deploy an Agentforce agent to answer warranty and parts availability questions  Track distributor portal adoption and case deflection rate
Retail	Integrate online and in-store purchase history into Data Cloud to power personalized email and SMS campaigns  Deploy a self-service AI agent for order status and returns  Track repeat purchase rate and net promoter score
Higher Education	Create an AI agent to answer FAQs (internally and externally)    Track answer helpfulness and knowledge article usage

 

The Fork in the Road 

Your organization faces a choice. Continue patching legacy automation, being reactive to every new release and every business pivot, or invest now in a composable, governed, AI-ready platform that scales with your ambitions. 

The cost of inaction is rising. Every quarter you delay modernizing architecture, data, and DevOps, you accrue technical debt that will block the next acquisition, product launch, or AI deployment.  

The cost of action is manageable if you sequence it correctly: start with quick wins (retire aging features, deploy one AI agent), build momentum with measurable KPIs, and reinvest savings into the next wave. 

The question is not whether your CRM will change. The question is whether you will control that change or be controlled by it. 

By: Corey Doohan