Mid-Market Technology Trends: From Salesforce to Sage, AI ROI Starts With the Foundation
Welcome to Mid-Market Technology Trends, a roundup of the technology and business developments most relevant to middle-market organizations.
Every issue, we cut through the noise to highlight the developments we believe are most likely to shape technology decisions, along with our perspective on what they mean for middle-market businesses.
Sage Is Telling Customers Where Its New Technology Will Land
At its partner conferences this month, Sage said it is simplifying its product portfolio around Sage Intacct and Sage X3. Those two products will get the newest technology and agentic AI first. Sage 100 and Sage 300 are not being abandoned. However, a Sage product leader said that putting new technology on the older platforms is difficult and expensive, and isn't a good investment for Sage, its partners, or its customers. Sage did not announce end-of-life dates or migration timelines.
Our Take: Vendor end-of-life stories rarely start with a date. They start with a statement about where investment is going. Sage 100 and 300 customers aren't being pushed to move. But the gap between what their system can do and what Intacct and X3 customers receive will widen with every release. The best time to plan an ERP migration is while the current system still works, not after that gap becomes a business problem.
What Happens When Your CRM Shows Up in Every App Your Team Uses?
Salesforce reported that usage of Agentforce Coworker grew 21 times in three months, to more than 185,000 users. Agentforce Coworker is an AI assistant that brings Salesforce data and agents into everyday work. This month, Salesforce is making Coworker generally available in Microsoft Teams, on iOS, and as a desktop app for Mac and Windows. It is also releasing an MCP server. MCP (Model Context Protocol) is an emerging standard that lets AI tools connect directly to business systems. One early customer, Fulton Bank, runs more than 20 Coworker use cases supporting about 3,000 users.
Our Take: AI in CRM is leaving the CRM screen. When sales and service teams can pull and update customer records from Teams or a phone, the quality of those records and the permissions behind them matter more than ever. The growth figures are Salesforce's own, and the company hasn't disclosed the starting baseline. Treat the trajectory as a signal, not a benchmark.
Only 4% of Service Leaders Have Seen the AI Return They Expected
A Gartner survey of 275 customer service and support leaders, conducted in July and August 2026, found that just 4% have achieved the ROI they expected from their AI initiatives. Gartner's guidance points to what separates the organizations making progress. They invest nearly 1.8 times more in data and knowledge foundations than in the AI use cases themselves. They pursue fewer use cases, each tied to a strategic objective. And they train staff on when and how to use AI tools.
Our Take: Customer service has been one of the most aggressive early adopters of AI, which makes it a useful early read on AI ROI for every other function. Returns depend less on which tool was purchased and more on whether the data and knowledge underneath it were ready. When measuring results, a good first question is how much of the budget went to the foundation versus the feature.
Your Data Platform Is Coming to Excel, Not the Other Way Around
Snowflake released CoCo for Excel and Google Sheets in public preview at its Expedition event on November 3. The tool lets finance and operations teams ask plain-language questions of Snowflake data from inside a spreadsheet, while the data itself stays in Snowflake. Snowflake also announced always-on AI agents, called Bots, that work in the background on recurring tasks. They check in with a person only when a decision or approval is needed. The Bots are slated for a limited private preview.
Our Take: Spreadsheet exports are where governed data usually stops being governed. Every download creates a copy that starts drifting from the source the moment it's saved. Bringing the data platform into Excel, rather than pulling data out of it, is a meaningful improvement in control. That holds only if the metric definitions underneath are agreed on first. The background agents are further out, but they show where data platforms are heading.
Security Responsibilities Are Outgrowing the Teams Assigned to Them
A new survey of 361 IT and security leaders at organizations with fewer than 5,000 employees found that 62% say limited staffing has left important security risks unaddressed. Nearly half (46%) say their most significant known security gap has been open for at least six months. Three in four say security responsibilities have grown faster than IT headcount over the past three years. The gap widens on smaller teams. Among IT teams with fewer than 15 people, 81% reported unaddressed risks, compared with 37% on teams of 15 to 50. The survey also found that 77% of organizations let employees use generative AI tools, but only 16% have comprehensive AI policies.
Our Take: Most of these gaps come down to capacity, not budget. Security tools bought without the time to configure and manage them end up as shelfware. The distance between allowing AI and governing it is the same pattern showing up across this issue. For lean IT teams, deciding which risks to own internally and which to hand to a managed provider is now a leadership decision, not only an IT one.
What This Means for the Middle Market
Across ERP, CRM, customer service, data, and cybersecurity, this cycle's developments point the same way. AI is moving into the places where work already happens, whether that's Teams, Excel, or the finance system. Vendors are also deciding where their best capabilities will live, and Sage made that explicit.
What sits underneath these tools determines the result. Gartner's data shows foundations separating the few organizations seeing returns from the many that aren't. The security research shows what happens when responsibility grows faster than the people who own it.
Ascendex helps organizations decide which platforms and capabilities deserve investment, carry those decisions through execution, and stay involved to make sure their systems keep working as the business grows.
Contact us if you're looking for help deciding what technology initiatives to prioritize and how to execute them with confidence.