“If you would not pitch them again today, they should not be on your client roster next year.”
Investors look for one thing first: does the revenue base get stronger or weaker over time. The same rule applies to your client list. The businesses that grow fastest in tech quietly fire the bottom 10% of their clients every year. They do it to protect margin, protect teams, and protect brand equity. The math is usually blunt: when founders drop their worst clients, profit per employee goes up, churn of good clients goes down, and enterprise value grows faster than topline alone would suggest.
The market does not reward vanity revenue. It rewards reliable, high-margin, low-drama revenue. That is why private equity firms almost never keep every client after an acquisition. They prune. They rank accounts by profit, payment behavior, expansion potential, and support load. Then they remove the bottom slice. Software founders talk about “net revenue retention” on pitch decks, but they rarely show the hidden lever: active curation of who they sell to. The top decile of SaaS companies treat their client list like a portfolio, not a museum.
The trend is not clear in every sector yet, but there is a growing pattern: operators who aggressively manage the bottom of their client base see more predictable growth and an easier time raising capital. Revenue quality becomes part of the story. Boards pressure CEOs to exit “toxic” customers. Product leaders request that sales stop selling fringe use cases that slow the roadmap. Support and success leaders quietly ask who they can stop serving. All of those threads tie back to a single question: which 10% of clients are pulling the company backward.
From a business value perspective, pruning the bottom 10% is about ROE: return on energy. Every sales call, feature request, and support ticket has a hidden cost in engineering time, brand risk, and opportunity loss. When your team spends its best hours on the wrong accounts, you pay for that twice. Once in direct cost. Again in what you did not build, sell, or improve for the right accounts. Investors do not just ask “how big is revenue” anymore. They ask “how expensive is this revenue to keep.”
“Founders fixate on closing deals. The best ones fixate on which deals they refuse to renew.”
Why the bottom 10% of clients are so expensive
Every company has three kinds of clients:
1. Profitable fans
2. Neutral, low-noise accounts
3. Loss-making, high-friction accounts
The first group funds your future. The second group is your growth buffer. The third group slowly erodes both.
The lowest 10% by value rarely look bad on a basic revenue report. Many pay on time. Some even look large. The problem sits in everything your P&L does not show at first glance.
They distort your product roadmap
A misfit client asks for edge-case features, custom work, and rushed integrations. Those requests sound attractive during a slow quarter. They come with statements like:
– “If you can ship this, we will expand.”
– “Our industry just needs one extra workflow.”
– “Legal needs this one custom clause in the SLA.”
Engineering then spends weeks or months on work that benefits a small fraction of clients. Product-market fit subtly drifts. That drift shows up later as:
– A cluttered interface that feels confusing
– Code paths that are hard to maintain
– Bugs triggered only by one or two heavy users
This feature debt drags your velocity. New investors run technical due diligence and ask why the roadmap is so fragmented. That hurts valuation multiples.
They create bad revenue quality
Not all revenue is equal. A $1 million client that churns with lawsuits and negative reviews is worth less than a $300k client that expands quietly for five years. Revenue quality shows up in:
– Margin
– Predictability
– Customer happiness
– Expansion rate
Bottom clients usually look weak on at least two of these. They:
– Demand discounts
– Threaten to leave unless they get custom work
– Delay payments
– Escalate minor issues
“Bad revenue hides inside gross ARR. Clean revenue shows up in net retention and margin stability.”
When you carry too much bad revenue, your gross figures look good, but your story sounds weak in the boardroom. Investors know that one or two large, low-quality accounts can compress valuation by an entire turn of ARR.
They burn out your best people
Top engineers, CSMs, and account managers rarely quit because of your best clients. They quit because of constant emergencies created by a small group of accounts that:
– Ignore onboarding
– Refuse standard processes
– Expect 24/7 access to senior staff
Every hour a senior engineer spends on a crisis call for a low-fit client is an hour not spent shipping features for the top 20% of clients. Over a year, that adds up to:
– Slower roadmap
– More production incidents
– Higher staff turnover
Internal morale drops. People start saying “we sell to the wrong clients” in private chats. That sentiment hurts productivity, hiring, and long-term retention.
The investor view: why pruning drives valuation
When investors review a tech business, they do not just ask “how many clients do you have.” They ask:
– What does the revenue curve look like by cohort
– How many accounts actually generate profit
– How noisy is support volume by segment
– Where is churn concentrated
The bottom 10% of clients usually appear in the red zone on all four questions.
Revenue quality metrics investors watch
Investors look for patterns like:
– High gross margin with stable or growing net revenue retention
– Low concentration of revenue in a small number of customers
– Predictable expansion in core segments
– Controlled churn that comes mostly from low-fit accounts
When founders show that churn is intentional and centered in the bottom tier, that signals discipline. It signals that the company protects its product and people from misfit business.
Imagine two SaaS companies, A and B, both at $10M ARR.
– Company A keeps every client. Gross margin 68%. Net revenue retention 96%. Support tickets per $100k ARR: 35.
– Company B trims the bottom 10% annually. Gross margin 78%. Net revenue retention 108%. Support tickets per $100k ARR: 18.
Both show growth, but investors will pay a different multiple. Company B earns a richer story: the business can grow without throwing bodies at support.
How pruning affects growth metrics
Here is a simplified view of how pruning the bottom 10% can change your growth picture over three years.
| Metric | Keep All Clients | Prune Bottom 10% Yearly |
|---|---|---|
| ARR after 3 years | $19M | $18M |
| Gross margin | 69% | 80% |
| Net revenue retention | 95% | 110% |
| Support FTEs | 40 | 25 |
| Valuation multiple (ARR) | 5x | 8x |
| Implied valuation | $95M | $144M |
The company that pruned has slightly lower ARR, but higher margin, stronger retention, and far better operational leverage. That turns into a higher multiple and a higher absolute valuation.
From a founder’s perspective, that is the core ROI: less revenue, more value.
How to identify your bottom 10% objectively
The worst way to fire clients is by gut feeling alone. The best way is through a simple scoring model that highlights real drag. This keeps the process fair and avoids emotional bias.
Build a client quality score
You can score each client across a few concrete dimensions:
– Profit contribution
– Payment reliability
– Expansion potential
– Support and success load
– Strategic fit
Then you rank clients by total score and review the bottom cohort carefully.
Here is a simple table to illustrate:
| Factor | Score Range | Measurement |
|---|---|---|
| Profit contribution | 1 – 5 | Net margin after support & infra costs |
| Payment reliability | 1 – 5 | On-time payments, disputes, credit risk |
| Expansion potential | 1 – 5 | History of upsell, product fit for more users |
| Support load | 1 – 5 (reverse) | Tickets, escalations, custom work requested |
| Strategic fit | 1 – 5 | Match with ICP, reference value, brand impact |
Support load can be inverted, so 1 means “extreme load” and 5 means “very light.” That keeps the total score intuitive.
After you score, the lowest 10% of scores usually stand out. They share traits:
– Below-average margin
– High ticket volume per dollar
– Irregular payments
– No expansion in 12 to 24 months
– Misfit with your ideal customer profile
This is the first draft list. Do not fire all of them blindly. Use judgment.
Listen to your team data
Alongside numeric scoring, gather qualitative data. Ask:
– Which clients create constant emergencies
– Which accounts require founders in every meeting
– Which clients threaten to leave every quarter
You can do a simple 3-question pulse survey for account managers and support:
1. “If you could stop working with three clients, who would they be.”
2. “Which clients request work far beyond what they pay for.”
3. “Which clients ignore our process and SLAs.”
Names that show up repeatedly often match the bottom 10% by score.
The right way to fire clients
Firing clients carries risk if done poorly. The goal is to protect revenue reputation while improving your internal economics. That requires clean communication and clear alternatives.
Set criteria before you act
Before any conversations, define clear exit criteria such as:
– Below X% gross margin for Y quarters
– More than Z escalations per quarter
– Repeated SLA violations or abusive behavior
– No product fit with current or future roadmap
Write these criteria down. Share them with leadership and legal. Point every exit conversation back to those criteria.
This protects you against accusations of arbitrary treatment and keeps the process consistent across segments and geographies.
Design an offboarding playbook
A clean offboarding playbook usually includes:
– Standard notice period based on contract
– Clear written explanation that focuses on fit, not blame
– Referral to partner agencies or platforms where possible
– Structured data export and handoff
– Defined last day of support and access
You can even treat it like a project with milestones. That calms both sides.
There are two common positioning angles:
1. Product focus: “We are narrowing our focus to clients in A, B, and C categories. Your needs sit outside that focus. We believe you will be better served by a provider focused on your use case.”
2. Commercial integrity: “To keep service levels high for all clients, we need commercial terms that reflect support needs. We cannot sustain the current model. Here are the options.”
In both, the message is straightforward: the fit is no longer there, and you are taking responsibility for acting on that reality.
Offer options, not negotiations
You can offer choices, but avoid open-ended haggling. Common options:
– New pricing that reflects actual usage and support
– Migration assistance to a partner or alternative solution
– Short-term extension to help them transition
Frame these as final options, not starting points. That prevents months of back-and-forth that drains even more energy from your team.
“The worst client exits drag on for quarters. The best ones finish in weeks, with clear dates and clean handoffs.”
How to avoid refilling the bottom 10% with new problems
Firing clients only works if your sales and marketing processes do not keep replacing them with similar ones. That is where ICP discipline and pricing strategy matter.
Refine your ideal client profile with real data
Every time you prune, update your ICP using actual performance data. Look at your best and worst clients in parallel. Ask:
– Which industries show the highest net margin and lowest noise
– Which company sizes adopt your product with the least friction
– Which buyer roles champion the product and renew without drama
Turn that into a sharper ICP. For example:
– From “mid-market e-commerce brands” to “DTC brands with 50k to 500k monthly visits, in fashion or beauty, with internal growth teams.”
Then enforce this ICP in:
– Lead qualification rules
– SDR scripts
– Marketing messaging and case studies
Your future client list becomes a mirror of what you accept at the top of the funnel.
Fix pricing that rewards unhealthy usage
Often, the bottom 10% exists because old pricing lets them consume huge amounts of support and infrastructure for a low fee. To fix this, map out how your pricing ties to cost drivers.
Build a simple table like this:
| Cost Driver | Current Pricing Link | Proposed Adjustment |
|---|---|---|
| Support tickets | Not linked | Introduce higher-tier plans with premium support fees |
| API calls / compute | Flat fee | Usage-based tiers for heavy users |
| Custom work | Included | Bill hourly or via scoped professional services |
| Onboarding hours | Included | Charge one-time onboarding for large or complex clients |
When pricing reflects cost drivers, misfit prospects either pay enough to justify the load or self-select out.
Managing internal resistance to firing revenue
Your sales team will push back. Your finance lead may worry about short-term numbers. Even co-founders may disagree. The way you frame “firing the bottom 10%” internally matters as much as the external messaging.
Present the math clearly
Build a simple internal memo that shows:
– Current revenue and margin from bottom 10%
– True cost of serving them (support, infra, discounts, write-offs)
– Modeled impact of removing them and reallocating effort to higher-value segments
For example:
– Bottom 10% revenue: $1.2M
– Direct costs: $600k
– Estimated overhead tied to these clients: $300k
– Net contribution: $300k
Then estimate the upside:
– If half of that freed capacity adds just $800k in new high-margin business at 80% gross margin, you add $640k gross profit.
– Even if you only reach part of that, the net gain is clear.
When people see that the bottom 10% can actually reduce total profit, resistance drops.
Protect sales compensation and morale
Sales teams hate losing accounts they fought for. To keep them engaged:
– Exclude fired accounts from clawbacks on commission, unless there was clear misrepresentation
– Share the scoring model so they understand why these accounts are off-fit
– Update sales targets to reflect more focus on ICP-fit deals, not just raw volume
You can even add a small incentive for deals from high-fit segments that later show strong margin and low support load. That links comp to quality, not just quantity.
Timing: when to prune and how often
An annual 10% cut sounds harsh, but in practice it becomes part of a routine rhythm. The key is to tie it to planning cycles, not emotional reactions.
Align pruning with planning cycles
A simple rhythm looks like:
– Q3: Run account scoring, identify candidates
– Early Q4: Finalize list, prepare communication, update forecast
– Late Q4: Begin offboarding with aligned timeline into early Q1
That gives you time to:
– Adjust revenue targets
– Update investor guidance
– Plan hiring based on a cleaner book of business
Some companies prefer a rolling approach, where they run the scoring quarterly and exit smaller chunks more often. The yearly 10% concept still holds as a target band.
Watch for signals that you waited too long
If these patterns appear, your pruning cadence is likely too slow:
– Support feels near constant crisis
– Product roadmap keeps bending around one or two accounts
– Leadership spends meaningful time on account-level disputes
– New hires learn early that “these few clients dictate everything”
Quicker action not only fixes current drag but also signals to the company that you protect focus.
The psychological hurdle: why founders avoid this move
None of this is complex at a spreadsheet level. The real barrier is emotional. Founders remember early clients that took a risk on them. Teams fear negative reviews. CFOs worry about covenants and growth targets.
Loyalty vs fit
There is a tension between loyalty to early adopters and the needs of your current stage. Early clients might have:
– Pushed your roadmap forward
– Given credibility when you had no brand
– Referred other accounts
At later stages, some of those same clients might now be:
– On legacy pricing far below cost
– Using outdated features
– Needing support on old versions
You can honor the history while still acting on current fit. That can look like:
– Grandfathering some features but moving them to partner support
– Offering migration credits to move them to a solution that matches their use case better
– Giving extended notice and hands-on help during offboarding
Loyalty does not require you to ignore the economics forever. It means exiting with care and respect.
Fear of public backlash
Some founders worry that fired clients will attack them online. That risk exists. To reduce it:
– Communicate in writing with clear, calm language
– Avoid blaming the customer; focus on focus and fit
– Give reasonable time and help for migration
– Keep a record of all interactions
Most tech buyers understand that vendors change direction. If you exit cleanly, the story in the market is rarely dramatic. Investors care more about your margin and retention than about a few loud LinkedIn posts.
Case pattern: how pruning often plays out over 24 months
Across many SaaS and tech-enabled businesses, a similar pattern appears once they start pruning the bottom tier annually.
First 3 to 6 months:
– Leadership stress rises slightly as they manage exits
– Revenue line dips or flattens
– Support and engineering feel immediate relief from the loudest accounts
6 to 12 months:
– Product velocity improves
– NPS from top and mid-tier clients rises
– Sales teams close more ICP-fit deals due to clearer positioning
12 to 24 months:
– Net revenue retention climbs
– Margins hold or improve even as revenue grows again
– Hiring plans focus more on product and growth, less on support backfill
From a valuation perspective, this two-year window is where the decision to fire the bottom 10% every year starts to pay back strongly.
When not to fire the bottom 10%
The rule is powerful, but not universal. There are edge cases where a strict 10% cut each year may not fit.
Some examples:
– Very early stage companies with fewer than 15 clients, where each account is a major share of revenue
– Highly concentrated enterprise plays where a small number of accounts pay eight figures and negotiation cycles are part of the model
– Regulated sectors where sudden offboarding could trigger compliance or legal complications
In those situations, the principle still helps: identify loss-making or misfit clients and decide what to do. The specific percentage and timing can vary.
The trend across later-stage tech is still clear: companies that treat their client roster like a portfolio, pruning the weakest positions to double down on winners, build stronger, more durable businesses.
“Growth is not just about adding revenue. It is about upgrading who you earn it from, year after year.”