Why Services-as-Software^™ Will Eat the SaaS Stack

For two decades, SaaS delivered a powerful trade: lower switching costs, faster deployment, predictable economics in exchange for processes that looked roughly the same at every company that bought the same product. That trade made sense when scale was the prize. It no longer does. The next decade of enterprise software belongs to a different model that HFS Research has named Services-as-Software(TM) and the leaders who internalize this will build durable advantages while their peers spend another budget cycle chasing parity.

I came to this conviction through a particular path. I started in strategy consulting, then moved into the business process re-engineering work at Deloitte during the years when that discipline was being invented. That vantage point, strategy at the top and process redesign in the wiring, taught me that a function's competitive advantage is almost never in the tooling. It is in the judgment of the leader running it and the accumulated tacit knowledge of the team they have built. When I later co-authored Selling the Cloud and then Data and Diagnostic-driven Selling, I was working out the same idea from different angles: SaaS adoption was about economics and scale, but selling differently than your competitors was always about extracting signals they could not see and acting on it faster than they could.

The SaaS Commoditization Problem

The math of SaaS rewards uniformity. A vendor's product roadmap has to serve the median customer in its market, or the unit economics break. That has been a feature for twenty years. It is now a bug for any function whose mandate is to differentiate. If your sales development team uses the same outbound platform, the same scoring model and calls the same intent signals as five of your direct competitors, you are not buying an advantage. You are buying parity. And parity, by definition, does not grow market share.

This is the awkward truth sitting underneath most enterprise tech stacks today. Boards bought SaaS to standardize. Now they are asking for the same functions that standardized to deliver differentiated growth with the same tools the competition has.

What Services-as-Software(TM) Actually Changes

The term Services-as-Software(TM) was coined by Phil Fersht and his team at HFS Research, and it captures something important about where enterprise work is headed. The unit of delivery is no longer the seat. It is the outcome. The product is not a workflow shipped to every customer in the category; it is an agent built around a specific company's data, a specific leader's playbook, and a specific operating context. The economics flip with it. You pay for the process, not the license.

What makes this powerful is not the technology itself. It is the working model that the technology enables. To build an agent that runs a function better than the SaaS alternative, you have to sit in the same room as the leader of that function, get into the details of how they actually make decisions, and encode the parts of their judgment that no software vendor will ever see because no software vendor sits in their offices. This is consulting work. Close to the team. Close to the leader. Deep in the specifics. It is the work I spent my career learning to do, and it is finally being given a scalable delivery mechanism.

Example: Lead Generation

Lead generation is the cleanest case. Every growth-stage company knows pipeline is existential, and every one of them has bought the same three or four platforms to do something about it. The result is a market in which the top of every funnel looks roughly the same: similar prospect lists, similar sequences, similar intent triggers, similar scoring logic. Reply rates fall every quarter for exactly this reason, not because the tools have gotten worse, but because everyone is running the same playbook on the same buyers.

A Services-as-Software(TM) approach starts somewhere else. We sit with the CRO and the demand-generation leader. We map the qualification heuristics that live in their heads, the buying-committee patterns they actually trust, the disqualifying signals their best AEs spot in week one, the territory-specific intent data their competitors do not even know to look for. Then we build an agent that runs the lead generation process around that judgment, not around a generic SaaS template. The output is not a faster version of the same playbook the competition is running. It is a playbook the competition cannot replicate because they do not have your CRO.

Beyond Sales: Where the Same Model Wins

Lead generation is the most visible case because pipeline is existential, but the mechanic, extract the leader's judgment, encode it, run it at software economics, is not specific to revenue functions. It works anywhere a function has accumulated know-how the competition cannot copy.

Accounts Payable and Financial Operations: A great controller is not valuable because they process invoices quickly. They are valuable because they know which suppliers tend to dispute, which invoice patterns hint at duplicate billing or fraud, which exceptions warrant escalation versus auto-approval, and which approvers in the business will hold up a month-end close if they are not nudged early. Generic AP automation handles the easy eighty percent and dumps the hard twenty back on the team. A Services-as-Software(TM) agent built around the controller's actual decision logic handles the hard twenty the way the controller would, and the close cycle shortens by days, not minutes.

Procurement and Supplier Risk: Every procurement organization runs the same source-to-pay platforms, yet outcomes vary wildly. The reason is rarely the tooling. It is the CPO and the category managers whose internal models for supplier reliability, built over years of late deliveries, quality slips, and recovery behavior, outperform any generic scorecard. Encoding that judgment into an agent that screens incoming bids, flags risk on existing contracts, and prioritizes which supplier reviews are worth a human's time turns a back-office function into a real source of working-capital and continuity advantage.

Talent Acquisition: Every company runs an ATS. Every ATS scores candidates roughly the same way. The recruiting leader who actually fills senior roles is not relying on that score; they are pattern-matching on signals their best hires shared. An agent trained on the recruiting leader's actual hiring history and the performance patterns of the team's top contributors can pre-screen at a quality level the generic platforms cannot approach, and it sharpens every quarter as more data accumulates.

IT Security Operations: A modern security stack generates thousands of alerts. The CISO and the senior analysts they trust are the only people in the company who reliably know which patterns are background noise in their environment and which signal genuine risk. That judgment is locked in their heads and walks out the door when they leave. Encoded into an agent, it scales across every shift, stays in the company when the analyst does not, and turns institutional security knowledge from a flight risk into an asset that compounds.

Finance, procurement, talent, security. None of these are revenue functions, and all of them are places where the same logic that works in lead generation works just as well. Any process where a leader's accumulated judgment is the moat is a candidate. Which, looked at honestly, is most of the processes that actually matter inside a company.

Why SaaS Could Not Do This and Services-as-Software(TM) Can

SaaS averaged that judgment away because it has to. The economics of the model demand it. Services-as-Software(TM) encodes the judgment because it can. That is the entire shift, stated plainly. And it applies to any process inside a company where doing the work a little differently, a little smarter, a little more specifically, is worth real money.

My View as an Advisor

I see this pattern reinforced in my advisory work with AGS, Genpact, and a top three global strategy firm, organizations whose operating models are themselves the competitive product. They have always understood that the value is in how the work gets done, not in which tools are used to do it. The shift HFS Research is naming gives the rest of the market a way to operate on the same principle, at software level economics, for the first time.

At Accelerant Growth Solutions, this is the model we are building around. Not buying more seats. Not deploying more dashboards. Sitting with the leaders who run the functions that decide whether a company grows, pulling out the tribal knowledge that has been trapped in their heads, and turning it into agents that run those functions better than any competitor running an off-the-shelf platform can match.

The next ten years of enterprise software will not be won by the companies that bought the most licenses. They will be won by the companies that turned the judgment of their best leaders into systems that compound. That is the promise of Services-as-Software(TM) and the leaders who move first will set the terms for everyone who follows.

Services-as-Software(TM) is a term coined and trademarked by HFS Research. AGS gratefully credits HFS Research and Phil Fersht for defining the category.