PSA tools track projects. They do not govern the chain

Most professional services automation tools solve two problems well: tracking projects and generating invoices. The recurring failure in professional services is upstream of both.

It sits in whether the terms you agreed at the point of sale exist as operational data the system can enforce, or as a document somebody has to read and apply by hand.

That is a category-level observation, not a competitive one. The tools in this market are mostly good at what they were designed for. The question is whether what they were designed for is the thing that is actually costing you money.

The mistaken diagnosis

Firms usually enter a PSA selection believing they have a visibility problem. Nobody can see project status, utilisation is a guess, and the reporting arrives too late to act on. So the evaluation becomes a comparison of dashboards.

Better dashboards do not fix it, because reporting is a view over records and the records disagree. If sales, delivery and finance each hold a defensible version of the same engagement, a dashboard does not reconcile them — it picks one and presents it with more confidence than it deserves.

The problem is not that you cannot see the number. It is that the number is not governed.

The five gaps

Where the category tends to stop, in our experience, is consistent enough to list.

Invoice lines that cannot be proven back to source. The invoice generates cleanly, but tracing a line back to the work, the approval and the contract term that permits billing it means opening three systems and asking two people.

No hard invoice-readiness gate. Nothing prevents an engagement reaching month-end with unapproved time, an undocumented variation and a milestone in an ambiguous state. So close becomes a rebuild rather than a release.

WIP as a report rather than a register. The balance is calculated from whatever is in the system at the moment you ask. A register is composed of entries that each hold their own justification, which is a different structure, not a better query.

Contract terms treated as documents. Rate exceptions, caps, discounts and change thresholds live in a PDF attached to an opportunity. A PDF cannot refuse an unbillable hour.

CRM, delivery and finance genuinely disconnected. Integrated at the level of records being copied between systems, which produces two versions that drift, rather than one record that several functions read.

None of those is a feature gap you would notice in a demo. Every one of them is a monthly cost you would notice in a close.

Bicycle or spaceship

The practical difficulty in this market is that the choice usually presents as two bad shapes.

At one end, tools that are simple, tidy and pleasant to use, but which cover part of the operating model — good project tracking, thin commercial governance. At the other, highly configurable enterprise platforms that will govern anything you like, and bring an implementation programme, a systems integrator and a budget line to match.

A firm of forty people with a real commercial-control problem falls between them. It needs the governance and cannot absorb the implementation, and the honest answer for a long time has been to buy the tidy tool and keep the spreadsheets.

That gap is the space we built into. The design constraint was governing the full chain — pipeline, scoping, delivery, time, costs, WIP, invoicing — without the firm having to become a software implementation project for a year. On Salesforce; works with your Salesforce org, or available as a bundled edition if you don’t use Salesforce.

When a lightweight tool is the right answer

This is the most important paragraph in the post, and vendors in this category habitually skip it.

A governed spine is the wrong purchase for a firm that does not have a commercial-control problem. A ten-person practice with one billing model, one rate card and monthly invoicing does not need contract terms held as enforceable data, because the terms fit in one person’s head and that person is in the room for every decision. Telling that firm it needs a spine is precisely the vendor error this post is criticising.

The threshold is not headcount, though headcount correlates. It is whether commercial truth has to travel between people who were not present when it was agreed. If it does not, buy the tidy tool, keep the overhead low, and revisit the question when it starts to.

There is a second limit worth stating. Governing the chain means agreeing billing rules, approval gates and change thresholds before delivery starts. That is real work by senior people, paid up front, and some firms will look at it clearly and decide they would rather keep the flexibility and absorb the argument later. That is a legitimate commercial position, not a failure of understanding.

What to evaluate instead of features

If you are running a selection now, the question that separates the category is not which tool has more capability. It is which one can answer three things without a person in the loop: what is this engagement allowed to bill for, what state is each of its commitments in, and can any invoice line be traced back to the work, approval and term behind it.

Ask that in the demo. Ask it about a messy engagement rather than a clean one — a variation mid-flight, a rate exception agreed in negotiation, a milestone accepted verbally. The answers will separate project tracking from commercial governance faster than any feature matrix.

How we hold that chain is set out on our invoicing and WIP page, how we sequence an implementation is on the quickstart page, and our commercial model is explained on the pricing page.