I want you to think about the last significant client your firm lost. Not a small piece of work that ran its course - a proper relationship. A client that had been with you for years, that accounted for real revenue, that you thought of as "one of ours."
Now think about when you found out. Was it a phone call? An email from their procurement team? A slightly awkward conversation at the end of a quarterly review? However it happened, I'd bet good money on one thing: by the time you heard about it, the decision had already been made. Weeks ago. Possibly months.
And I'd bet something else. If you went back and looked - really looked - at the six months before that conversation, you'd find a trail of signals so obvious in hindsight that you'd feel a bit sick. Shorter emails. A meeting cancelled, then another. A portal they used to log into weekly that they hadn't touched since February. A newsletter they'd been opening for three years that suddenly started bouncing off their inbox unopened.
You weren't looking for those signals. Almost nobody is. And that's the thing I want to talk about.
Clients who leave professional services firms rarely do so in a blaze of anger. The dramatic "we're taking our business elsewhere" phone call makes for a good story, but in my experience it accounts for maybe one in ten departures. The other nine follow a pattern so consistent it's almost boring.
It starts with small frictions. Nothing dramatic - a deliverable that was fine but not quite what they expected. A response time that slipped from same-day to end-of-week. A junior team member handling something that used to get partner attention. Individually, none of these are worth raising. They're too small, too petty-sounding, too much effort to articulate. So the client absorbs them.
Then comes the withdrawal. The client starts investing less in the relationship. Meetings get shorter. Briefings become less detailed - they stop giving you the context you need to do your best work, which ironically means your work gets worse, which confirms whatever narrative is forming in their head. They're not hostile. They're just... less present.
Then - and this is the bit that really stings - they start looking. Quietly. A conversation with a competitor at a conference. A look at another firm's website. A reference check with a peer who uses someone else. This happens entirely out of your view. By the time they've had two or three of these conversations, they've already built a mental case for switching. The formal notice is a formality.
I watched this play out with a mid-market consulting firm we worked with last year. They lost a client worth roughly £400k in annual fees. I was sitting in a fairly grim meeting room in their offices, the kind with a whiteboard that's never quite clean, when I asked the relationship partner what happened. He said, "They just decided to go in a different direction." He was holding a coffee he hadn't touched. When I pushed him on whether there had been any warning signs, he went quiet for a moment. "Looking back," he said, "they'd been pulling away for about six months. I just thought they were busy."
They weren't busy. They were leaving. They'd just started leaving before anyone noticed.
I've done the same thing, if I'm honest. Years ago, a client we'd worked with for three years went quiet on us. Slower replies, a couple of check-in calls that got rescheduled and never rebooked. I told myself they had a lot on. They left six weeks later. I knew, somewhere. I just didn't want to look at it directly.
We know our important clients. If they were unhappy, they'd tell us.
I hear this constantly. And I understand why you believe it - you've built relationships over years, you've had honest conversations, your door is always open. But here's the thing: complaining requires a specific kind of emotional investment that a disengaging client has already withdrawn.
Think about it from their side. To give you useful feedback about what's going wrong, they'd need to sit down, articulate the problem clearly, explain what they'd want to be different, and then wait to see if you actually change. That's work. Quite a lot of work, emotionally. And it requires something that's already eroding: belief that you're capable of changing the thing that's bothering them.
A client who's experienced a pattern of small frictions - the slow response times, the junior-heavy team, the slightly generic advice - has already formed a view about how the firm operates. They're not thinking "I should raise this so they can fix it." They're thinking "this is just how they are now." And once someone reaches that conclusion, departure becomes easier than feedback.
This is the bit that's genuinely painful to accept. Switching to another firm involves hassle, sure - new onboarding, new relationships, a period of reduced efficiency. But it's a one-time hassle with a defined endpoint. Complaining, waiting, hoping, being disappointed again? That's open-ended. Most people will choose the thing with a clear finish line.
I've spoken to a few clients over the years - after they'd left a firm, not ours, thankfully - and the language is remarkably consistent. "It wasn't one thing." "I didn't want to make it into a big deal." "I just felt like we'd outgrown them." None of them described a dramatic failure. All of them described a slow realisation that the relationship wasn't giving them what it used to.
Right, so here's where this gets practical. Because the disengagement pattern I've just described isn't invisible. It's just not being monitored.
If you have a client portal - and most professional services firms of any size do - you're sitting on one of the clearest leading indicators of client health you could ask for. Login frequency. That's it. Not sophisticated analytics, not AI-powered sentiment analysis. Just: how often is this client logging in, and has that changed?
A client who was logging into your portal weekly to check case updates, download reports, or review billing, and is now logging in once a month, has told you something important. They've reduced their investment in your service infrastructure. They're spending less time inside your ecosystem. That's a signal, and it's sitting right there in your usage data - probably being looked at by nobody.
Same goes for email engagement. If you're sending newsletters, thought leadership, or sector updates - and you should be - your email platform already tracks open rates at a contact level. A client who opened every email for two years and has now ignored the last four hasn't just "got busy." They've mentally checked out of your content relationship. Your marketing team has this data. I'd wager nobody is cross-referencing it with your client revenue list.
Then there's response time. This one's harder to track systematically, but it's perhaps the most telling signal of all. When a client who used to respond to your emails within hours starts taking three days, something has shifted. They're deprioritising you. Not because they're rude - because their internal calculus about the value of the relationship has changed.
And finally - the canary in the coal mine - event attendance and relationship call availability. If your top client has declined two consecutive quarterly reviews, that's not a scheduling conflict. A client who values the relationship makes time for it. A client who's already halfway out the door finds reasons not to.
I've written separately about what client satisfaction surveys miss - the companion piece to this one goes deeper into why behavioural data tells you things that survey responses can't. But the short version is this: surveys measure what clients say they feel. Behavioural signals measure what clients actually do. When the two diverge, the behaviour is always more reliable.
Here's where I think most firms get stuck. They hear "monitor client engagement signals" and immediately picture some kind of expensive data platform or a six-month analytics project. And then, because that sounds expensive and complicated, they do nothing.
It doesn't need to be complicated. Honestly, you can set up meaningful early warning monitoring with tools you already have. Three things that any firm with a portal, an email marketing platform, and a CRM can implement in a week.
A monthly portal engagement export. Pull login data for your top clients. Sort by trend. You're looking for clients whose login frequency has declined for two consecutive months. That's your watch list. A spreadsheet that someone updates once a month is fine - the point isn't real-time monitoring, it's pattern detection.
A quarterly email engagement review by client. Take your email marketing data and segment it by client account. Identify the bottom 20% by open rate. Cross-reference that with your revenue list. Any top-revenue client sitting in the bottom 20% for email engagement should trigger a proactive relationship conversation. Not a sales call - a genuine check-in.
A communication response time log for your top twenty clients. This is the most manual of the three, but also the most valuable. Once a month, have someone review the last few email exchanges with each of your top twenty clients by revenue. Flag any where average response times have lengthened materially compared to three or six months ago. I know this sounds tedious. It is a bit tedious. But it takes about two hours and it might save you a £300k client.
None of this requires specialist data infrastructure or a technology investment. It requires someone to do it, which means it requires someone to decide it matters. That's the real barrier. Not capability - priority.
Worth saying here: portal engagement isn't just an early warning signal, it's also a retention tool in its own right. A portal that clients actually want to use creates stickiness that makes switching harder. A portal they've stopped using has already removed one of your retention mechanisms.
In my experience - and I want to be clear this is based on patterns I've observed across engagements rather than a peer-reviewed study - the intervention window between the first significant disengagement signal and the point of no return is roughly four to six weeks. After that, the client has typically done enough competitor research, had enough internal conversations, and built enough momentum towards a switch that reversing the trajectory becomes extremely difficult.
Four to six weeks. That's not a lot of time. But it's enough, if you're watching for it.
The firms I've seen handle this well share a common trait: they treat the signal as a trigger for curiosity, not panic. They don't call the client in a cold sweat saying "we've noticed you haven't logged into the portal!" That would be weird and a bit creepy, frankly.
What they do instead takes some subtlety. The relationship call comes first - but not a generic "just checking in" call. Something specific: "We've been reviewing how our clients are using the portal and I wanted to get your perspective on whether it's working for you the way it should." That's an invitation to give feedback without making it feel confrontational. It gives the client permission to raise the small frictions they've been absorbing. And sometimes - not always, but sometimes - that conversation reopens a door that was quietly closing.
Then, proactive content delivery. Not the newsletter they've stopped reading - something specific to their situation. A note from a partner saying "I saw this and thought of your business" with a genuinely relevant article or insight attached. The point isn't the content. The point is demonstrating that someone at the firm is still actively thinking about their specific challenges. It's a counter-signal to the "they've stopped caring about us" narrative the client may be building.
The one that firms most often get wrong is senior partner involvement. If a significant client is disengaging, and the signals suggest they feel they've dropped down the firm's priority list, the intervention needs to come from someone whose involvement signals otherwise. Not the associate who manages the day-to-day. The partner. Or the managing partner. A lunch. A call. Something that says, without saying it, "you matter to this firm at the highest level."
All of this has to happen before the client has started actively evaluating competitors. Before they've had the internal conversation with their board about switching. Before the decision has crystallised. Because once it has, you're not intervening in a disengagement - you're negotiating a departure. And those are very different conversations.
I want to step back for a moment, because there's a broader point here that I think gets missed. The signals I've described - declining portal usage, reduced email engagement, longer response times - are symptoms. They're the visible manifestation of a growing gap between the experience your client expects and the experience your firm is delivering.
I've written about this experience gap in more detail elsewhere, and it's worth reading if this piece has made you uncomfortable, because it gets at what's driving the disengagement patterns I've been describing. But the short version is this: your clients' expectations aren't being set by your competitors. They're being set by every digital experience they have - the banking app that remembers their preferences, the SaaS platform that surfaces relevant content before they search for it, the e-commerce site that makes returns effortless.
When your portal requires three logins and a password reset to access a document, they're not comparing that to other law firm portals. They're comparing it to Monzo.
That comparison might not be fair. But it's the one they're making.
Let me make this concrete. Take a mid-market professional services firm - 200 fee earners, say - with a top-twenty client list averaging £250k in annual fees each. That's £5m in revenue from your most important relationships. If you lose two of those clients a year because you didn't spot the disengagement signals - and based on what I've seen, two is conservative for firms that aren't monitoring - that's £500k in annual revenue gone. Not immediately replaceable, either, because the cost of acquiring a new client at that revenue level is significantly higher than the cost of keeping an existing one.
Now compare that to the cost of the three monitoring approaches I described. A monthly portal export. A quarterly email review. A response time check on twenty accounts. Maybe half a day of someone's time per month. Call it a full day to be generous. That's twelve days a year to protect £5m in revenue.
The maths isn't subtle.
Do this today, or at worst this week. Pull up your portal analytics. Look at your top ten clients by revenue. Check their login frequency over the last six months.
If the trend is flat or increasing, brilliant. Sleep well.
If it's declining for two or more of them, you have a problem that hasn't announced itself yet. And you have a window - maybe four to six weeks, maybe less - to do something about it before the only conversation left to have is "we've decided to go in a different direction."
That conversation, as you might already know, is one of the worst in professional services. Not because of the revenue - though that hurts - but because of the nagging feeling that you could have seen it coming. That the signals were there. That someone, somewhere, had the data.
They probably did. The question is whether anyone was looking at it.
If you want to understand which of your current clients are showing the early warning signals described in this piece - and what the intervention approach would be for each - book a client retention risk assessment. Sometimes the most valuable thing an outside perspective can do is point at the thing you've been too close to see.