Domain expertise has long been a differentiator for services firms. The ability to go deep in a particular industry or functional domain not only creates credibility and differentiation, it also improves sales efficiency and makes delivery more repeatable. Knowing how work gets done, which decisions matter, unwritten rules and necessary tradeoffs, can be the difference between a highly successful program outcome. Or a complete failure.
That expertise has always mattered, but in the AI era, it matters more than ever before. It’s why we created the Domain Leaders category in this year’s Tercera 30 research.
Why is AI Making Domain Expertise More Important?
Once AI moves from assisting people to performing work, domain expertise brings three things vital to success: prioritization, context and judgment.
Prioritization. AI is currently being touted as the answer to every problem, yet ROI remains elusive for most. When companies need to make a case for, or defend, AI spend, it’s important to know which use cases will create the most value versus what is just easy to automate.
Context: Context is critical to workflow redesign. Domain experts are uniquely situated to account for the dependencies, trade-offs, and exceptions that generic solutions can miss. They also understand what data lives in existing platforms, what it’s used for, how it relates to other data, and when it can create new revenue opportunities.
Judgement. This piece is perhaps most important. Having the knowledge and reps across similar problems can make it a whole lot easier to evaluate an AI solution’s performance. That judgement can shape more realistic test cases, help decide where human review matters most, and quickly recognize plausible but wrong answers. And then what to do about it.
Consider this example: an insurer using AI to support claims handling. A claims expert helps identify which evidence is missing, which exceptions need investigation, and when a case should be escalated. They know how to test whether AI is making sound recommendations. Whether it’s using realistic scenarios involving incomplete documentation, conflicting evidence, or unusual policy terms. They recognize output quality quickly, and gauge what’s happening behind the scenes. They will know whether faster processing is creating value or more risk.
The same principle applies within business functions. A call center expert understands why reducing average handling time can backfire if customers have to call again. An HR specialist knows that a skills model must reflect the work people actually do. A finance expert understands which reconciliation exceptions are routine and which warrant investigation.
With AI, domain expertise can become reusable IP
Historically, much of this knowledge lived in people’s heads. An experienced supply-chain consultant knew how to handle an inventory exception. A finance specialist understood how to account for an unusual transaction. A contact center leader knew what circumstances meant departing from the script.
AI now makes it possible to capture elements of that knowledge in agents, workflows, prompts, policies, and evaluation frameworks. Instead of applying an expert’s judgment one engagement at a time, a services firm can turn parts of it into reusable assets.
BCG’s study of AI front-runner companies describes the role of a “domain anchor.” Someone who can encode expertise into reusable form.
However, capturing expertise is only the beginning. Once knowledge is encoded, some of it becomes easier for others to access or replicate. The durable advantage comes from testing it, adapting it to each client, and improving it as conditions change. All of this presents an ongoing service opportunity.
Small language models are another way to make domain expertise reusable. In the insurance example above, a services firm might fine-tune a small model on expert-labeled claims to identify missing documentation or categorize exceptions. It could retrieve the applicable policy language, and craft explicit workflow rules to determine which cases require human review. The resulting IP could include the model, the agent, training examples, knowledge base, rules and evaluations.
Gartner predicts that specialized, domain-specific agents will generate 80% of tangible agentic AI ROI by 2028.
What does this mean for software partnerships?
This all begs the question: if AI makes software and domain specific workflows easier than ever to create, why continue investing in partnerships with software vendors?
It’s true that some enterprises may choose to replace or forgo buying applications or features with software or workflows they build themselves. Services firms may even play a role in helping to evaluate this decision.
However, the ability to create software or a workflow does not eliminate the need for reliable records, access controls, transaction processing, and ongoing maintenance. Enterprise-grade platforms provide these foundations, along with years of embedded business logic. Their value extends well beyond the interface to the data, controls, and dependencies that keep a business running.
Replacing them means migrating historical data, rebuilding integrations, validating controls, and changing how people work. AI can reduce the effort of writing replacement software, but the enterprise still has to establish that the new system can operate reliably. And then, take responsibility for evolving and maintaining it.
Vertical and horizontal systems of record, such as those listed in our Tercera 30 research, can provide excellent foundations for AI. Agents need trusted information, clear permissions, and dependable ways to execute actions and record results. Established platforms can supply those capabilities, while services firms build specialized workflows within and across them.
The opportunity is to understand which foundations are worth retaining, where custom capabilities create an advantage, and how to connect both around the client’s desired outcome.
Where the Tercera 30 fits
This growing importance of domain expertise in the AI era inspired our new Domain Leaders category in the Tercera 30. For this year’s report, we selected 10 pillar platforms that span industries and functional domains, including financial services, healthcare, engineering, talent, and go-to-market (just to name a few).
These partner-forward, AI-relevant vendors create meaningful opportunities for firms that specialize in these markets. In most cases, domain specialists will work across multiple vendor partners. However, these pillar platforms provide a solid place to start.
To learn more about the Domain Leaders and other software ecosystems creating opportunities for technology services firms, explore the full Tercera 30 report out on Oct 15.