Segmentation: How to Allocate a Customer Portfolio by Potential
Written by Alex Raymond, founder of AMplify and author of The Growth Department.
Customer segmentation in Account Management is the practice of scoring every account in the portfolio and assigning it to a tier that decides how much time, attention, and investment it receives. Most companies sort by revenue alone. Joanna Hagelberger grew an account paying $50,000 a year into one worth $10 million in annual revenue, an account a revenue-only model would have handed to nobody. This guide covers the four scoring factors, the tiers, allocation, cadence by tier, and the annual reset.
Why segmentation is part of the Post-Sale operating system
Segmentation is the first artifact of the Commitment phase of the Growth Department Method. Clarity settles what Post-Sale is for and who owns its revenue. Commitment turns that agreement into three working artifacts: the Segmentation Plan, account plans for the top 20% of accounts, and the Risk Register. Segmentation comes first because the other two depend on it. You cannot decide which accounts deserve a full account plan, or how much risk review a renewal needs, until you have decided which accounts matter most.
The economics make the case. Frank Cespedes of Harvard Business School has observed that the old 80/20 rule has tightened: in many modern portfolios, roughly 10% of customers produce roughly 90% of returns. On the other end of the curve, 60% to 70% of accounts will renew flat no matter what the team does, and only 3% to 5% will produce expansion that changes the company's year. A portfolio shaped like that punishes equal treatment.
Most teams know this in principle and still spread effort evenly in practice. The loudest account gets the most hours. The friendliest contact gets the most meetings. The account that pays the most gets the senior Account Manager, whether or not it has anywhere left to grow. By the way, your Chief Financial Officer already sees the portfolio this way. Every budget cycle, the finance team asks which customers are worth the cost of serving them. Segmentation is how the Post-Sale leader answers that question first, with a model the executive team can inspect, before someone else answers it with a spreadsheet and a cut.
Segmentation also protects the team. When every account is treated as a priority, Account Managers hold portfolios they cannot serve, and the strategic work gets squeezed out by whatever is urgent this afternoon. A written Segmentation Plan forces trade-offs into the open, where leadership has to agree to them, so no Account Manager is left making them alone at 6 p.m.
What is customer segmentation in Account Management?
A Segmentation Plan is the written, leadership-approved model that scores every account, assigns it to a tier, and defines what service, coverage, and cadence each tier receives. It has three parts: the scoring model, the tier definitions, and the operating rules for each tier.
The plan answers four questions. Who are our customers, account by account? Which of them have the most potential to grow? How will we serve each group differently? Who on the team is assigned to which group? A team that can answer all four in writing has a Segmentation Plan. A team that answers with "our enterprise reps take the big ones" has a habit.
The distinction between segmentation and risk management matters too. Segmentation decides how much investment an account deserves. The Risk Register tracks what could cause the account to churn. The two artifacts work together, and they answer different questions. A Strategic-tier account can sit on the Risk Register, and a Scale-tier account can be perfectly healthy.
Three signs tell you the Segmentation Plan is missing or out of date. First, nobody can say how many accounts sit in each tier without opening a spreadsheet that someone built last year. Second, the same Account Manager covers a top-ten customer and dozens of small ones, and the small ones get answered first because their requests are quick. Third, the expansion forecast comes from the largest accounts by default, and when the Chief Revenue Officer asks where next year's growth will come from, the answer is a list of the biggest logos. Each of these is a sign that revenue alone is deciding where the team spends its time.
The fix starts with an honest conversation about trade-offs. A Segmentation Plan only works if it changes behavior: some accounts get more than they get today, and some get less. If the plan changes nothing about who serves whom and how often, the team has produced a label. The test for any segmentation model is simple: does it force a trade-off the team would rather avoid? If it does, it is doing its job.
Why sorting by ARR fails
The default segmentation model in most companies is a single line drawn through annual recurring revenue. Everyone above $50,000 gets a named Account Manager. Everyone below gets a pooled team or a digital program. It is easy to build, easy to explain, and wrong in two directions.
Joanna Hagelberger, who built two Account Management functions from scratch in insurance technology, calls this the chalk line. "Everybody below 50,000 in revenue doesn't get an account manager, and that's the gut reaction," she said in an AMplify Building the Growth Department session. "But what you lose in that is you could have a very small company revenue-wise that's paying you 10,000 dollars a year that's worth 2,000,000. Because you drew a chalk line and said they're not important because they're not paying us a lot today, you're missing the potential opportunity."
Her own career is the proof. At Vertafore, she took a division of a global insurance company from $50,000 to $10 million in annual revenue by asking a question nobody had asked them: where do you want to be in five years? At her next company she grew a $200,000 account to $5 million. Both accounts started below the line most teams would have drawn.
The chalk line fails in the other direction as well. Hagelberger points to large, famous companies that build their own technology and have no intention of buying more from a vendor. "Just being big does not make them high potential," she said. "They have to be willing to partner as well." A revenue-only model gives those accounts your best people and your most expensive coverage, year after year, with no expansion to show for it.
Justin Strackany, former Chief Customer Officer at SecureLink, put the same problem in one line: if FedEx and a local mechanic shop are both paying $10,000, they are not the same account. Revenue tells you where the portfolio stands today. It says almost nothing about where the portfolio is going.
The four scoring factors
AMplify's Segmentation Model scores every account on four factors and combines them into a single composite score. The four scoring factors are Current Revenue, Potential, Fit, and Commitment.
Current Revenue is what the account is worth today: trailing twelve-month annual recurring revenue, or annualized monthly revenue. It is scored 1 to 5 by portfolio position, so the top 20% of accounts by revenue score a 5, the next 20% score a 4, and so on.
Potential is the realistic expansion revenue the account could produce over the next 24 months, measured in dollars. Dollars work better than multiples, because a 10x multiple is easy to write on a $100,000 account and hard to defend on a $5 million one. Dollar estimates also add up to a portfolio total the Chief Revenue Officer can take to the board: identified expansion potential across the customer base. Potential is also scored 1 to 5 by portfolio position. Hagelberger's advice applies here: find the markers in your industry that show an account is small with you and large in the world.
Fit is a direct 1 to 5 rating of how well the account matches the company: ideal customer profile alignment, use case match, vertical relevance, and technical fit. An account with poor fit costs more to serve and rarely expands, however large it is.
Commitment is a direct 1 to 5 rating of how invested the customer is in the partnership: executive sponsorship, mutual investment, and partnership behavior. This is the factor that catches the famous logo that will never buy more. It also rewards the smaller customer whose leadership treats you as part of their plan.
The default weights are 30% Current Revenue, 30% Potential, 20% Fit, and 20% Commitment. Revenue and Potential are scaled against the rest of the portfolio because absolute dollars mean different things at different companies; a $500,000 account is enormous for one company and a rounding error for another. Fit and Commitment are rated against an absolute standard, because strong executive sponsorship looks the same in any portfolio. All four end up on the same 1 to 5 scale, so the composite score works.
The model produces two rankings side by side: rank by revenue alone and rank by composite score. The accounts that rise under the composite model are the bets the team should be discussing. The accounts that fall are usually over-served, and the conversation there is about right-sizing investment while keeping the customer well served. The model surfaces these differences. The team decides what to do about them, with the Chief Revenue Officer and Chief Financial Officer in the room.
Allocating accounts by tier
The composite score sorts every account into one of three tiers, with a fourth tier above them for concentration.
Strategic accounts score 4.0 or higher. They get a named Account Manager, a full account plan, and executive engagement. These are the foundation accounts, where retention and expansion both matter and the full discipline of the Method applies.
Standard accounts score between 3.0 and 3.9. They get pooled coverage, a structured cadence, and expansion plays run from a playbook. They are healthy revenue served efficiently.
Scale accounts score below 3.0. They are digital-first by design, with human intervention at trigger moments such as a usage drop, a renewal decline, or an expansion signal.
Mega accounts sit above the model. Any single account worth more than 10% of total annual recurring revenue is designated Mega automatically, because concentration risk at that level is a C-suite problem. Mega accounts get a named C-suite sponsor, custom coverage, and their own cost-to-serve.
Strackany ran a version of this at SecureLink, which grew from $1 million to $12 million in annual recurring revenue during his time leading sales, and later reached $50 million. New customers were scored at onboarding, with adjustments for potential size, and placed into tech-touch, managed, strategic, or executive tiers. The executive tier held five accounts, each with an internal executive involved in every major decision.
Two allocation rules matter more than the exact thresholds.
One Account Manager works one tier. Mixing tiers in one portfolio is one of the most common mistakes AMplify sees. An Account Manager with two strategic accounts, four mid-sized accounts, and seventy-five small accounts will spend the week on whatever is loudest, and the strategic accounts get the leftover hours. Dedicated people or teams for each tier allow each group to build work rhythms that fit the accounts they serve.
Account plans go to the top 20% only. Strategic and Mega accounts get individual account plans. Standard and Scale accounts run from playbooks. This concentrates planning effort where it produces the most return, and it keeps account planning from turning into a compliance exercise nobody reads.
The model gives a starting range for portfolio size: one account per Account Manager at the Mega tier, two to five at Strategic, six to 25 at Standard, and 26 or more at Scale, where the Account Manager runs triage on top of digital engagement. Company maturity and contract size decide where in each range a team lands. If the staffing math does not work (80 Strategic accounts and capacity for 50, for example), tighten the tier definition or reduce the number of tiers. An under-resourced tier performs worse than a smaller, properly staffed one.
Cadence follows the tier
Segmentation decides who gets which meetings. Each tier has its own operating rhythm, and the rhythm should be written down alongside the tier definitions.
At the Strategic tier, contact is biweekly to monthly and multi-threaded across the customer organization, with a quarterly business review, one or two on-site visits a year, and a VP-level sponsor who joins key meetings. Renewal forecasting starts nine months before the renewal date. At the Standard tier, a monthly check-in with the primary contact and a lighter semi-annual business review are usually enough, and renewal forecasting starts six months out. At the Scale tier, touchpoints are triggered by signals in the data, and renewals run from a standard motion. Mega accounts get weekly to biweekly contact across several stakeholders and renewal planning that starts twelve months or more ahead, with the Chief Revenue Officer involved. The renewals guide covers the renewal side in detail.
Kristy Devantier, Managing Director at TaleWind Digital, runs a clear version of this. Her team holds Strategic Review Meetings with clients four times, twice, or once a year, depending on each client's engagement and growth potential. The highest-potential clients get the most time with the team, and the company president joins monthly internal account reviews. Her results: a 100% annual recurring revenue renewal rate and more than 8% revenue growth. The cadence works because the segmentation decided in advance which clients earn which rhythm.
The Annual Segmentation and Strategy session
Segmentation is a living model, and it needs two clocks.
The quarterly clock handles promotion and demotion. An account goes up a tier when its expansion pipeline crosses a defined threshold or its Commitment score rises by two points or more. An account goes down a tier when revenue declines by 20%, or when Commitment drops by two points or more for two consecutive quarters. These rules keep the model current between rebuilds, so the tiers reflect the portfolio as it stands now.
The annual clock is the Annual Segmentation and Strategy session, one of the seven meetings that run a Growth Department. Once a year, a broad group of leaders looks across the entire portfolio and decides how to divide it, what each tier means, how accounts are allocated to each Account Manager, and what the goals are for each segment. The weights and thresholds get re-debated. The question of whether to segment by industry, geography, or product gets reopened. Hagelberger's former company, for example, later split its Account Managers by market as well as tier, with property and casualty insurance on one side and life and health on the other, so each Account Manager could become an expert in the customer's world.
The annual session is non-negotiable. Segmentation built once and never revisited drifts out of date within a year, and the tiers start describing a portfolio that no longer exists.
Putting it together
Read Hagelberger's two accounts through this model. A $50,000 account and a $200,000 account would both score low on Current Revenue. Both would score high on Potential once someone looked for the markers of size in the world, and high on Commitment once the customer's leadership started planning with her. Under a composite model, both rise into the Strategic tier and get a named Account Manager, a real account plan, and executive attention. Under a chalk line, both sit in a pooled queue. The difference between those two outcomes, in her case, was more than $14 million in annual revenue.
That is what segmentation is for: putting your best people in front of the accounts most likely to reward them, and doing it by design.
Where to start
Pull your full customer list with current revenue this week. Add a realistic 24-month expansion estimate for each account, then rate Fit and Commitment from 1 to 5. Rank the list twice, once by revenue alone and once by the composite score, and look hard at the accounts that rise and fall. Take both rankings to your Chief Revenue Officer and Chief Financial Officer, agree on the tiers, and assign each Account Manager to a single tier. Then put the Annual Segmentation and Strategy session on next year's calendar before anything else claims the date. The full installation sequence, from Clarity through Cadence, is in the Post-Sale operating system guide.
The Growth Department is the post-sale revenue operating standard for B2B companies. Take the ten-minute Post-Sale Stress Test to see where your function stands across Clarity, Commitment, and Cadence. Or download the free audiobook of The Growth Department and read the manifesto.
About the author Alex Raymond is the founder of AMplify and the author of The Growth Department, the operating standard for the function that delivers most of a company's revenue. He spent a decade building Account Management and Customer Success systems with B2B companies before founding AMplify, where he works with Post-Sale leaders on installing the Growth Department Method. He hosts the Account Management Secrets podcast.
