> ## Content Index > Fetch the complete content index at: https://www.revenueoperationsalliance.com/llms.txt > Use this file to discover other available public pages before exploring further. # Account scoring should explain not just rank - URL: https://www.revenueoperationsalliance.com/account-scores-reps-actually-use/ - Published: 2026-10-13T17:06:57.000Z - Updated: 2026-10-13T17:14:04.000Z - Description: How to build an account score reps can read and act on, calibrated against your closed-won deals and reused to find strong-fit accounts outside your CRM. - Author: Anthony Goldbloom - Tags: RevOps strategy, Account-based marketing, Lead scoring, Articles ****TLDR: The best account scores do more than rank accounts. They show reps where to start and what to say.** - Show reps the rank and the linked evidence, not just the raw score. - Break the score into segments you can read (e.g. size and growth/momentum). - Avoid large accounts winning by default by using growth and concentration as attributes. - Be wary of attributes that don't cover the whole corpus: funding data only exists for venture-backed companies, so scoring it artificially boosts startups. - Calibrate your scores against closed-won accounts (or a subset of accounts that are a strong ICP fit). - Use the same model to find whitespace accounts: strong-fit companies not yet in your CRM. Most account scores get ignored. They're either a proxy for company size or a black-box number. Reps tune out the first because it says nothing new, and the second because they can't see why an account got an 82. A score earns its place when a rep can read it, trust it, and act on it. A good score points the rep at a first move, not a grade. Every number Sumble provides is backed by the people, teams, and projects behind it. **Get the skill:** Use this free skill to turn your ICP into a scored, clickable prospecting queue. Works in Claude Code, Codex, or Cursor. 1. Copy*npx skills add SumbleData/agents --skill sumble-account-scoring* 2. Then start a new session and run */sumble-account-scoring*, or ask your agent to “use the sumble-account-scoring skill.” 3. Prefer to install by hand? [Download the repo ZIP](https://github.com/SumbleData/agents/archive/refs/heads/main.zip) and copy the *sumble-account-scoring* folder into your agent’s skills directory. 4. No npx or git? In Claude, add the plugin marketplace *SumbleData/agents* and install *sumble-account-scoring* from there. ## Why it's worth getting right Done well, an account score ends up underneath a surprising number of go-to-market decisions: - **Account allocation & territory planning**: assign every account to a rep and carve balanced territories from a ranked, scored universe — not by gut or geography alone. - **Rep guidance**: tell each rep which of their accounts show the most promise right now and why. - **Whitespace**: point to the same model at the accounts you don't have to find the net-new ones worth pursuing. - **Account-based marketing**: focus ABM spend and air cover on the accounts the model says matter. - **Lead routing**: when a lead comes in, its account score helps decide who works it and how fast. - **Disqualification**: filter out accounts that aren't an ICP fit. Get the score right once and it lifts the rest of your GTM engine. The rest of this guide is the method, which we've also encoded into a skill that builds it for you. ## No black-box scores: reps should be able to inspect and act on the attributes ![](https://storage.ghost.io/c/62/19/6219f8f5-5d35-43bc-82c3-ce0aeb6b2cc9/content/images/2026/10/Screenshot-2026-10-01-at-9.55.24---AM.png) The same Walmart account scored two ways: a points table versus a ranked score with linked drivers for size, growth and concentration Bias towards attributes that reps can understand and take action on. Some examples: - **People**: how many of your target personas work there, what share of the company they make up, and whether that function is growing. - **Technology**: how many of their teams use competitive and complementary technologies. - **Active projects**: relevant migration or transformation work that indicates an open buying window. [Know your accounts like you work there. | SumbleFind the teams, technologies, and active initiatives inside your target accounts - so you reach the right people with the right message.![](https://storage.ghost.io/c/62/19/6219f8f5-5d35-43bc-82c3-ce0aeb6b2cc9/content/images/icon/logo512-6b75c135-4c74-4a47-8875-9478f667d51b.png)Sumble![](https://storage.ghost.io/c/62/19/6219f8f5-5d35-43bc-82c3-ce0aeb6b2cc9/content/images/thumbnail/hero-mobile-4d606dfb-e2f1-4329-82e2-e30588be9265.png)](https://sumble.com/) ## Avoid having your model purely proxy company size Enterprises have more of everything, so a model built on raw counts is a headcount ranking in disguise. A few fixes keep size from dominating. If your sweet spot is small but fast-growing, weigh the growth attributes heavily, above all growth in your ICP personas. Fast-growing companies cross new scale points often, and each scale point is a moment they outgrow a tool and go looking for a new one, which makes them likely buyers of a modern solution. Lean on growth metrics here and the right accounts pop to the top: a company like Anthropic (\~4K employees) can be every bit as compelling as a Walmart (2MM employees). On the flip side, a company that isn't growing often isn't in pain because it isn't hitting the limits that trigger a change, so a big, static account can be a worse bet than a smaller one accelerating into your category. If what matters is that your solution is central to the business, weigh concentration: a high share of your ICP persona or of teams running the relevant technologies. If you sell to DevOps engineers, a company where 5% of headcount is a DevOps engineer is a far better fit than one where it's 0.05%: your product sits at the core of how they operate, not off in a corner. We also recommend a formula that keeps large outlier numbers from dominating the score; more on that in the scoring math appendix below. ## Be wary of attributes that don't cover the whole corpus The best attributes are ones every account could have a value for. If an attribute only exists for one slice of your universe, the model may end up overweighting membership in that slice. Funding is the classic example. Funding data only exists for venture-backed companies: bootstrapped businesses, PE-owned companies, subsidiaries, and public companies all read as zero. Score "total raised" and every venture-backed startup gets an artificial boost, while everyone else is penalized for a data gap, not a fit gap. And what funding is meant to indicate (fresh budget, an imminent hiring ramp) often shows up in attributes that do cover everyone, such as headcount growth. ## Break the score into segments you can read "Is this a good account?" hides several different questions, and the score is far more legible when its top level breaks them apart. We default to three segments, each answering one question, blended into a single number you can still take apart: - **Size**: how big is the opportunity? Persona headcount, number of teams using competitive or complementary tech, recent project + technology hiring. - **Growth & momentum**: is now the time? Growth in your ICP personas. - **Concentration**: how strong is the fit? The size-neutral ratios: what share of the company is your ICP persona, and what share of teams run your tech. Reps get one number to act on; anyone who asks can see the three lenses behind it, weighted \~50/30/20 (size / growth & momentum / concentration). That's a default starting point; the model calibrates it to your won deals, and you can adjust it yourself. The segments are yours to redefine: rename them, reweight them, or cut them a different way entirely. The most useful alternative is a business-unit breakdown: if you sell distinct product lines, give each its own segment (for Oracle, an OCI-fit segment and an Apps-fit segment, each with its own personas and technologies) and read a per-line score inside the same model. ## Calibrate against the deals you've won Weight factors so the scoring fits your closed-won accounts. If you do better with fast-growing companies, give growth in key personas a bigger weight. If you sell mostly to AI-native or digitally-native companies, boost them. If IT services shops are partners rather than buyers, penalize that category. The same gold set powers an evaluation view: tune the weights and watch whether your known customers rise to the top. ## Start with one model, not a model per market It's tempting to build separate scores for SMB, mid-market, and enterprise. Resist it. One model across all accounts is far easier to reason about and maintain, and it puts all your effort into getting a single thing right instead of keeping five things in sync. With the right features, especially size-neutral ones like concentration and growth, a single model that holds across the whole range is usually achievable. (Segments are different: those are lenses inside one model, and even a per-product-line breakdown is still a single score.) Build separate models only once you've proven one can't do the job. ## Show a rank, not a raw score, and rank within territories A raw score is challenging to interpret. Is 82 good? A rank is instantly legible: this is your #3 account. Treat the score as the engine and the rank as the interface: compute the score, then hand reps an ordered list. One catch: a single global ranking can feel unfair, because a handful of reps own all the top accounts and everyone else feels stuck with leftovers. The fix is to rank within territories: SMB, Enterprise, or Strategic, or by geography. This groups the ranking, not the model. One model still scores every account; you just order the results within each territory or size band. ## Re-run your scoring daily Most accounts barely move from week to week. But periodically a new migration project kicks off or hiring surges in a key persona, and an individual account moves dramatically. If you see these movements before your competitors, that can make you the front-runner. And if scoring is a daily job, it becomes one less thing to worry about. A score you rebuild by hand every quarter is a chore that rots between runs: someone has to remember to do it, the data drifts simultaneously, and you're never quite sure the list a rep is working is current. ## Use the same model to find whitespace accounts The model you calibrate against your CRM is really a description of what your best customers look like, and nothing about that description limits it to accounts you already have. Turn it around and the same weights surface the companies that look like your customers but aren't in your CRM yet. That's whitespace, and it comes almost for free. A company you've never engaged has no first-party attributes to join, so whitespace runs the Sumble-only half of your model (same normalization, same calibration multipliers, weights re-normalized over what's left) across Sumble's whole universe, then subtracts the accounts you already own. This is why it beats a bought "lookalike" list. A purchased list is firmographics, the company-at-rest description this method warns against. Whitespace is built on your own won deals and on what companies are doing right now, and it's the same model you already watched to sort your known customers to the top. Each account it surfaces lands with the people and the reasons to reach out already attached. ## The score is the start, not the end To recap: observable attributes, normalized so size doesn't dominate, calibrated against your wins, and every component clickable through to the people, teams, and jobs a rep can act on today. That's the difference between a score that sits in a dashboard and one that drives outreach. ## A step-by-step guide to world-class account scoring in under two hours The sections above argued that a good account score is the start of a prospecting conversation, not a verdict. Every number should be backed by people and teams a rep can see and act on. Here is how to build one against your own ICP using the [Sumble account scoring skill](https://github.com/SumbleData/agents/tree/main/skills/sumble-account-scoring), which runs in Claude Code, Codex, or Cursor. One skill, three objectives: score the accounts in your CRM, find whitespace (strong-fit orgs you don't yet have), or both in one ranked sheet. You talk to it, it pulls the data, it calibrates to your closed-won, and you tune with sliders. If python app.py runs, you are done. ### What you need - One coding agent: Claude Code, OpenAI Codex, or Cursor. - The Sumble account scoring skill, installed in step 1 above with one npx skills command or by copying a folder. - A Sumble account, API key, and the Sumble MCP connected. - Your CRM accounts if you want to score your own book: name and domain, provided as a CSV or pulled from a connected MCP/API source. The skill can use your whole CRM, which accounts are allocated to a rep, and your closed-won customers, as one table with flag columns or as separate lists. None are required, but closed-won is the one worth digging up: it powers the evaluation that checks your known-good accounts rise to the top. ### 1\. Install the skill A skill is just a folder of instructions your coding agent reads. There are a few ways to get it. If you have npx (it ships with [Node.js](https://nodejs.org)), one command installs the skill into whichever agents you choose: npx skills add SumbleData/agents --skill sumble-account-scoring The installer detects your coding agents and asks where to install. For a no-prompt global install into one agent, add -g -a claude-code -y, -g -a codex -y, or the matching agent name. No npx or git? Install it by hand: 1. Download the skills repo as a ZIP (no git needed): [github.com/SumbleData/agents → Code → Download ZIP](https://github.com/SumbleData/agents/archive/refs/heads/main.zip), then unzip it. 2. Copy the skills/sumble-account-scoring folder into your agent's skills directory (\~ is your home folder; create the directory if it doesn't exist): - **Claude Code:** \~/.claude/skills/sumble-account-scoring - **OpenAI Codex:** \~/.codex/skills/sumble-account-scoring - **Cursor:** \~/.cursor/skills/sumble-account-scoring No npx or git? Claude installs it as a plugin instead: **Customize → Plugins → Personal → "+" → Add marketplace → Add from a repository**, enter [SumbleData/agents](https://github.com/SumbleData/agents/tree/main/skills/sumble-account-scoring), and install sumble-account-scoring. Turn on **Sync** in the marketplace's "..." menu and later versions arrive on their own. Whichever route you take, start a **new agent session** so the skill is picked up, then run it: in Claude Code type /sumble-account-scoring; in Codex, Cursor, or the Claude desktop app, ask the agent to use the sumble-account-scoring skill. ### 2\. The interview The first step is short and scripted. The skill pulls a first draft of your ICP from your Sumble profile: personas, technologies, projects, and related buying attributes. You edit it in plain English until it matches how your team sells. It also proposes the score's **segments**, the top-level lenses it breaks the number into. The default is three: **Size** (how big the opportunity is), **Growth & momentum** (whether now is the time), and **Concentration** (how strong the fit is), with a starting blend of \~50/30/20\. That blend is a starting point, not the final weighting that the skill later fits to your closed-won deals (step 4). Keep them, reweight them, or redefine them; the skill suggests a business-unit breakdown if you sell distinct product lines. Funding deliberately isn't part of the default model: funding data only exists for venture-backed companies, so scoring it artificially boosts them (see above). If webinar attendance, whitepaper downloads, events, free-tier users, or product usage matter, point the skill at first-party data keyed by account and those become weighted factors too. ### 3\. Which accounts to score Every account lands in one category: customer (closed-won), allocated (in CRM, owned by a rep), unallocated (in CRM, no owner), or whitespace (a strong-fit org not in your CRM). The app shows that category as a column and lets you filter on it, so one ranked list serves territory planning, allocation gaps, and net-new at once. - **Score your accounts.** Point the skill at your account list and choose: score them all, or have it draw a representative **stratified sample** (≈30% closed-won, 40% rep-allocated, 30% unallocated). For a large CRM, where most accounts are stale and scoring everything just burns credits, a sample (or a subset you specify, like just rep-allocated) calibrates on the accounts that matter. It shows the credit estimate before any large pull. - **Fold in whitespace.** Say yes and the same skill, run in whitespace mode, ranks strong-fit orgs that aren't in your CRM (resolving your whole CRM universe to exclude what you already have) and drops them into the same sheet tagged whitespace. The ranking itself is free (only the final enriched pool costs credits), drawn from a diversified preselection so the pool isn't just the biggest companies. Candidates whose parent is already a CRM account are flagged as land-and-expand. Whitespace runs the Sumble-only half of your model, and the same gold-set calibration carries straight over. The skill shows a credit estimate before any large pull, then builds the app. ### **4\. Tune and click through** cd account\_scoring/ python app.py # http://localhost:8001 Drag the sliders and the ranking re-sorts instantly. Open any account for a per-attribute breakdown. This is where the score becomes actionable: each attribute deep-links into Sumble, filtered to the entities behind it. **DevOps engineer headcount** opens the actual engineers. **Teams using Jenkins** opens those teams. **Migration intent** opens the recent job posts. A rep goes from score to named contacts and talking points in two clicks. If you loaded CRM lists, the **category chips** above the table filter to customers, allocated, unallocated, or whitespace in one click, so you can pull up "high-scoring accounts with no rep" or "top whitespace" without exporting anything. The slider weights apply across every category at once. Use the sliders to correct for the bias you want to avoid. By default scores tilt toward big companies; the sliders are split so you can correct that on purpose. If you sell into small, fast-growing companies, turn growth up, especially ICP-persona growth. If you want accounts where your product is central rather than incidental, turn concentration up. Then use the Evaluation tab to see whether your won deals rise. The skill also fits to your wins carefully. It learns tag multipliers from over-represented and under-represented Sumble org tags among your gold set (for example a digital-native boost, or a professional-services penalty): calibration runs over org tags, not a per-industry multiplier; industries you obviously don't sell to are excluded or penalized in the interview instead. It makes a first guess at the weights, then a small solver takes a deliberately light pass, nudging only the high-level segment and category weights toward your closed-won accounts. It stays coarse to avoid overfitting: it never tunes at the most granular level (individual job functions, technologies, and projects keep their default within-segment weighting), shrinks changes back toward the defaults, cross-validates on held-out wins, and is adopted only when it generalizes. On a thin gold set it does not move at all. The fitted weights are a starting point you can override. The Evaluation tab shows the payoff: it reports how many gold accounts land near the top (a lift above 1.0 beats random) and breaks down where each category lands in the overall ranking, so you can confirm your customers cluster near the top. Drag, watch the gold rise, then Save: it writes your tuned weights and one complete score.csv. ### 5\. Score your whole book and keep it fresh Two artifacts come out of tuning: - **account-scoring-weights.json.** Your model as plain JSON: weights, normalization parameters, and each attribute's mapping back to Sumble data. - **score\_accounts.py.** A zero-dependency runner that reads the JSON and scores any list of accounts. For a one-off score across your whole book, run the portable scorer against Sumble's public API: python score\_accounts.py --accounts my\_full\_list.csv --out scored.csv For daily scoring, do not keep paying per API call. The lighter path is to have Sumble feed the underlying attributes into your warehouse or CRM on a daily refresh, then recompute the score on a schedule and write the rank back where reps see it. Teams with a data stack can re-implement the formula in SQL and run it through dbt, Airflow, Dagster, or Prefect. Teams with a little cloud infrastructure can use Cloud Scheduler and Cloud Run, or EventBridge and Lambda. GTM-owned teams can run the same workflow through Clay, n8n, Make, or Zapier. What you hand reps is the rank, not the raw score. "Your #3 account" beats "an 82" every time. Rank within segments or territories so every rep has a clear top of their own book. ## The payoff Say you sell a platform-engineering tool to DevOps and infrastructure teams. Within an afternoon, you can see whether customers skew digital-native and recently funded, whether Kubernetes footprint predicts fit better than headcount, and whether a recent raise plus an active cloud-migration initiative is your strongest "buy now" combination. Because every attribute links straight into the people, teams, and jobs behind it, the ranked list is a prospecting queue with the next action already attached. And the same skill, run in whitespace mode, turns it on the accounts you're not selling to yet: same model, minus the first-party attributes a stranger can't have. **Use the Sumble skill to get set up.** Our [account scoring skill file](https://github.com/SumbleData/agents/tree/main/skills/sumble-account-scoring) gives you the primitives to map and score your TAM. ## See it live Sumble is the AI-native data platform with the deepest insights on your TAM. Accurate and traceable intel on accounts, teams, contacts, technographics and more delivers foundational context for agentic workflows across revops, sales and marketing. Explore the [account scoring calibration demo](https://account-scoring-demo.sumble.com/): a sample book scored with sliders to retune the weights and per-attribute deep links on every account, the same kind of app the skill builds.