---
title: Account fit ROI kit
version: "1.0"
date: 2026-10-10
source: https://freshcontext.ai/our-work/builds/gtm-realignment
repo: https://github.com/Fresh-Context/roi-kits
license: CC BY 4.0
made-with: Fresh Context
---

# Account fit ROI kit

This file is a model of what it's worth when every team works the accounts that fit you best, written so Claude can build it for you. Claude reads it and makes two things: an interactive ROI calculator you fill in with your own numbers, and a one-page case for the budget you can send to finance.

It judges no vendor. The program cost is a number you type, whether it's a firm's quote or an in-house build. Every default is an example.

- **In claude.ai:** drop this file into a chat and say "build my ROI calculator".
- **In Claude Code:** say "read account-fit-roi.md and build the calculator".

The calculator carries a small "Made with Fresh Context" stamp. Ask Claude to take it off or put your own brand on, and it will.

---

## Instructions for Claude

The person who gave you this file wants an ROI calculator for account fit and coordinated GTM, and a one-page case for the budget. Everything you need is below: the conversation, the build, the model, the brand and two images. Follow it as written. Where the person asks for something different, do what they ask.

### 1. Open the conversation

Greet them in one line, then offer two ways to start:

1. **Use the example numbers** and change them in the calculator as you go.
2. **Answer three quick questions first**, one per row.

Then wait for their pick. If their first message already decides (they gave numbers, or said to use the examples), skip the offer and build.

If they pick the questions, send all three in one message and say they can skip any:

- **Win rate:** How many opportunities a year do you work on your best-fit accounts, and what's your average deal size?
- **Time:** About how many sellers and marketers work accounts, how many hours a week does each spend on accounts that don't buy, and what does an hour cost, loaded?
- **Pipeline:** How much pipeline did your best-fit accounts produce last year, and what's your win rate today?

Ask nothing else. The lift in win rate, the share of hours that comes back, the lift in pipeline, weeks worked and the program cost are judgment calls or near-constants. They're easier to set with the number moving in front of you, so they stay as fields at their example values. Don't ask for a company name, a team size or an industry. Never ask a follow-up for anything the calculator shows as a field.

### 2. Build the calculator

Make one self-contained HTML file: plain HTML, CSS and JavaScript, no framework, no library, no build step. In claude.ai, make it an HTML artifact. In Claude Code, or anywhere without artifacts, write it to `account-fit-roi.html` in the working directory and give the person the path.

No network calls except the Google Fonts stylesheet. No analytics, no CDN scripts, no fetched images, no browser storage. State lives in the page.

**Header.** The honeycomb tile from ASSETS at 56px square with a 14px radius, top-aligned with the title "Account fit ROI". Under the title, one line that says what the page is and teaches its one convention, for example: "What working your best-fit accounts as one team is worth, in dollars a year. A number tagged *example* is a placeholder, not research; the tag goes when you put in yours."

Under the header, two tabs: **Calculator** and **The case for the budget**.

**The three row cards,** in this order: Win rate on best-fit accounts, Time back from poor-fit accounts, Pipeline from coordinated plays. Each card holds:

- The row's label as the card heading.
- The row's value: the card's big number, typed directly into a field whose own text is Crimson Pro 600 at about 36px (style the input itself, so a general input rule can't shrink it). Its visible label is "Dollars a year". Typing a value overrides the inputs: tag the row *typed* (where the *example* tag sits), hide its aside, and show a "Recompute from the inputs" button.
- The row's aside, small under the value.
- The row's **what** text, verbatim.
- **Work it out**, a small heading, always open: the row's **how** text, verbatim, then the row's inputs, each with its label and its hint (verbatim, only where the model gives one). Changing any input recomputes the row and writes its value, replacing a typed one.
- A "Show the formula" toggle that shows the formula with the live numbers in it, for example `240 × $48,000 × 2 pts = $230,400`. For a typed row it reads "Typed: $X" and shows what the inputs would give.
- The row's low, base and high (see Sensitivity), small, in one line.

**The summary card.** The annual total, the sum of the three rows, in ink at about 56px with a 4px orange rule under it: that rule is the page's one highlight. Under it, the low and high totals and the stacked bar. Then an optional field, "What the program costs, first year", with the placeholder "Optional". It starts empty, with no example, and 0 counts as empty. When it has a value, show payback in weeks and first-year return, at base and at the low case. If a case's total is 0 or less, it reads "Doesn't pay back on these numbers" for that case. A "Show the formula" toggle here too.

**The stacked bar.** One horizontal bar, three segments sized to the three base row values, in row order. Label it with a legend under the bar, one line per row: a swatch, the row's short label and its value. Fill the segments `#381800`, `#8a7763` and `#c8b795`, with a 2px white gap between them. It's a data mark drawn from the numbers: plain divs or SVG rects, nothing decorative. If a row is zero or negative, leave it out of the bar and say so in one line under it.

**Sensitivity.** Three judgment inputs drive the range: "Lift in win rate, in points", "Share of those hours that comes back" and "Lift in pipeline from coordinated plays". Low halves each of the three. High multiplies each by 1.5, capped at 100. Every other input stays put. Say so on the page in one line: "Low halves the win-rate lift, the share of hours and the pipeline lift. High adds half to each." A typed row has no inputs to vary, so its low, base and high are all the typed value; say so on that row.

**Layout.** Content max width 1120px, centered, 16px side gutters on a phone, no sideways scroll at any width. Above 1000px, two columns, `grid-template-columns: minmax(0, 1fr) 360px`: the three row cards stacked in the wide left column, the summary card alone in the narrow right column, sticky at the top, so the total stays in view while they work. In the markup, the rows' wrapper comes first and the summary second. Below 1000px, everything stacks and the summary card moves to the top with `order: -1`. Inside a row card, set the inputs two to a line when there's room.

**Fields.** Let people type the way they talk: accept `48000`, `48,000`, `$48k`, `6m`, `25`, `25%` and `2 pts`. Reformat on blur: dollars as `$48,000`, percents as `25%`, points as `2 pts`, hours as `4 h`, counts bare. The unit lives in the field, never as a word beside the label. Arrow up and down step by the input's step; a row value steps by 1,000 and the program cost by 10,000. Clamp percents and points to 0 to 100 and everything else to 0 or more. Recompute on every keystroke. Round each row to whole dollars before you add them, so the total always equals the rows on screen. Every field has a real `<label>`, and the whole page works from the keyboard.

**The example tag.** Every input starts tagged *example*, and the tag goes when the person changes it. A row's value carries the tag while all its inputs are examples and the row isn't typed. Keep the tag quiet: small sans caps in `#8a7763`, beside the field's label.

**Check it before you hand it over.** With every example in place, the calculator must show:

| | Low | Base | High |
|---|---:|---:|---:|
| Win rate on best-fit accounts | $115,200 | $230,400 | $345,600 |
| Time back from poor-fit accounts | $103,500 | $207,000 | $310,500 |
| Pipeline from coordinated plays | $60,000 | $120,000 | $180,000 |
| Annual total | $278,700 | $557,400 | $836,100 |

The win-rate row's aside reads "4.8 more deals won a year." The time row's aside reads "2,300 hours a year, about 1.3 people's time." The pipeline row's aside reads "$480,000 more pipeline a year." With a program cost of $100,000, payback is 10 weeks at base and 19 at the low case, and first-year return is 457% at base and 179% at the low case. If any number is off, fix the code before you show it.

After the calculator, say two or three lines in the chat at most: which two inputs to look at first (the most sensitive, see section 3), and that you can write the case as plain text for an email.

### 3. The case for the budget

The second tab is one printable page for a finance partner. It reads the calculator's current numbers live. Write it plain and specific: no adjectives about the opportunity, no hype, no campaign lines. Lay it out as one column about 720px wide on the white ground, with hairline rules between sections, not cards. Each section opens on a plain heading in Crimson Pro 600, about 22px, sentence case. No uppercase or small-caps labels over sections. The Print button sits at the top right of the case.

In this order:

1. **Heading:** "Account fit: the case for the budget", and today's date.
2. **The number,** in one or two sentences: "On these assumptions, working our best-fit accounts as one team is worth $557,400 a year: $230,400 from a higher win rate on best-fit accounts, $207,000 in team time back from poor-fit accounts and $120,000 in bookings from coordinated plays. The range is $278,700 to $836,100." If a program cost is entered, add: "The program costs $X in the first year. It pays back in N weeks and returns R% in year one; at the low case, N weeks and R%." While any example remains, add one quiet line: "Inputs marked Example are illustrations, not our numbers."
3. **The assumptions,** as a table: every input, grouped by row, with its value and its source, "Example" or "Ours". Add the program cost as the last line, source "Ours", when it's entered. A typed row says "Typed as $X" in its group heading and keeps its inputs listed. Under the table, the formulas, one line per row.
4. **What would make this wrong.** Vary each of the three judgment inputs on its own, low to high, with everything else at base, and rank them by how far the annual total moves. Name the top two, one sentence each, in this shape: "**Lift in win rate, in points.** At 1 point instead of 2, the total falls by $115,200." With the examples, the second is "**Share of those hours that comes back.** At 25% instead of 50%, the total falls by $103,500." Skip a judgment input whose row is typed; if all three rows are typed, say the rows were typed in directly and the inputs aren't in play. Close with one sentence: the other inputs are facts, so check them against last year's actuals.
5. **The next test to run.** One test, aimed at the most sensitive input from step 4, with what to measure, how long it runs and the result that confirms or breaks the estimate. Use the matching template with the live numbers in it:
   - **Lift in win rate, in points:** "Score the open opportunities on one segment's best-fit accounts and put every team on that one list, beside a comparable segment worked the way you work today. Compare win rates once a full sales cycle of those deals has closed. The estimate holds at [lift] points or more. At [lift ÷ 2] points or less, plan on the low case."
   - **Share of those hours that comes back:** "For four weeks, have five sellers and marketers log the hours they spend on accounts the score rates a poor fit. For the next four, route those accounts away from them. The estimate holds if [back]% or more of those hours move to best-fit accounts. At [back ÷ 2]% or less, plan on the low case."
   - **Lift in pipeline from coordinated plays:** "Run one play with marketing, sales and customer success on the same best-fit accounts, beside a comparable set of accounts each team works alone. Compare pipeline created per account over a quarter. The estimate holds at a [lift]% relative lift or more. At [lift ÷ 2]% or less, plan on the low case."
6. **Footer:** "Model: account fit ROI kit, version 1.0." and the stamp (section 5).

The Print button calls `window.print()`. The print stylesheet prints the case from either tab, hides the header, tabs, calculator and buttons, keeps the 720px column, and fits on one page: 14mm margins and slightly smaller type will do it. If printing doesn't work where they are, they can ask you for the case as plain text to paste into an email. Write it when they ask.

### 4. Brand

Make it look like Fresh Context by default:

- **Ground:** white `#FFFFFF`, set on `body`, so it looks the same inside a dark chat.
- **Cards:** cream `#FBF8F0`, radius 22px, no border, and the shadow does the work: `0 1px 2px rgba(56,24,0,.05), 0 8px 24px rgba(56,24,0,.09)`.
- **Ink:** headings and values `#381800`; body text `#5b4636`; hints, tags and quiet text `#8a7763`; rules and field borders `#d8cfbc`. Never pure black, never a cold gray.
- **Accent:** orange `#F89420`, used as a mark and only as a mark: the rule under the annual total, links in the text, the focus ring, the active tab's underline. Never a large fill, never a button background, never body text.
- **Type:** Crimson Pro for headings, values and reading text. Merriweather Sans for labels, tabs, buttons, tags, units and hints. Load both from Google Fonts (Crimson Pro 400, 600, 700 and 400 italic; Merriweather Sans 400 and 700), with fallbacks `Georgia, serif` and `"Helvetica Neue", Arial, sans-serif`. Set `font-variant-numeric: tabular-nums` on numbers.
- **Fields:** white, 1px `#d8cfbc` border, radius 10px, a 2px orange focus ring.
- **Buttons:** Merriweather Sans, ink text on cream or white, 1px `#d8cfbc` border, radius 999px.
- **Images:** the honeycomb tile and the Fresh Context mark, as the data URIs in ASSETS. The honeycomb is AI Accelerated GTM's motif. The mark is 1.76:1, never square: set its height and leave its width auto. Don't recolor, crop, filter or redraw either one. Give every `<img>` an `onerror` that hides it, so a bad copy shows nothing rather than a broken icon. Alt text: empty for the honeycomb, "Fresh Context" for the mark. Never draw a stand-in.

### 5. The stamp

End both tabs with one quiet line: the mark at 24px high, then "Made with Fresh Context", the two together linking to https://freshcontext.ai in a new tab. Merriweather Sans, 12px, `#8a7763`; the stamp stays quiet, so it doesn't take the accent color links use.

If the person asks you to remove the stamp, remove it. If they ask you to rebrand the calculator to their company, do it. Either way, do it without comment. If they give you their colors or logo, use them: their main brand color takes the accent's job, their darkest color takes the ink's, and the ground stays light enough to read. A rebrand drops the honeycomb and our mark unless they ask to keep them.

### 6. Be straight with the numbers

- Show every formula.
- Label example numbers as examples, in the calculator and in the case.
- Never invent a benchmark or cite a statistic the person didn't give you. No "industry average", no "teams typically see". If they ask what's typical, say this kit carries no benchmarks and its examples are illustrations, not research.
- If a number they give looks off by an order of magnitude (a 40-point lift in win rate, 30 hours a week on poor-fit accounts, a $950 hour, $6 billion of pipeline), ask once. Then use theirs.
- Keep the model as written: three rows, these inputs, these formulas. The rows don't double count: the first prices a better win rate on the opportunities you already have, and the third prices new pipeline at today's win rate. If they want a fourth row or a different formula, build it, mark it as theirs and show its formula too.
- Make no recommendation about who should run the program. The program cost is theirs to fill in: a firm's quote or the cost of building it in house.

### 7. The words

- Plain operator English. Contractions. Short sentences. Name the actor and the thing.
- GTM, never spelled out.
- No em dashes anywhere. Use a hyphen with spaces, or rewrite the sentence.
- No hype words: not "unlock", "supercharge", "seamless", "game-changer", "revolutionize", and never "leverage" as a verb.
- Use the row labels, the what and how text, the hints and the asides exactly as the model gives them.
- Two lines of Fresh Context language may appear in the calculator, never in the case, each at most once: "Every team on the accounts you're built to win." and "Fit decides the win." They fit best in italic Crimson Pro: the first as the calculator's last line above the stamp, the second under the win-rate row's what text. Use one, both or neither.

---

## The model

Three rows, each in dollars a year. Every default is an example.

### Inputs

| Row | Input | Key | Unit | Example | Step | Hint |
|---|---|---|---|---:|---:|---|
| Win rate on best-fit accounts | Opportunities a year on best-fit accounts | `opps` | n | 240 | 10 | |
| Win rate on best-fit accounts | Average deal size | `deal` | $ | 48,000 | 1,000 | |
| Win rate on best-fit accounts | Lift in win rate, in points | `lift` | pt | 2 | 0.5 | Points, not percent: 2 turns a 25% win rate into 27%. |
| Time back from poor-fit accounts | Sellers and marketers who work accounts | `people` | n | 25 | 1 | |
| Time back from poor-fit accounts | Hours a week each on poor-fit accounts | `hours` | h | 4 | 0.5 | Research, sequences, meetings and follow-up on accounts that don't buy. |
| Time back from poor-fit accounts | Share of those hours that comes back | `back` | % | 50 | 5 | |
| Time back from poor-fit accounts | Loaded cost of an hour | `rate` | $ | 90 | 5 | |
| Time back from poor-fit accounts | Weeks worked a year | `weeks` | n | 46 | 1 | |
| Pipeline from coordinated plays | Pipeline a year on best-fit accounts | `pipe` | $ | 6,000,000 | 100,000 | |
| Pipeline from coordinated plays | Lift in pipeline from coordinated plays | `plift` | % | 8 | 1 | Relative lift. 8% turns $6M of pipeline into $6.48M. |
| Pipeline from coordinated plays | Win rate today | `win` | % | 25 | 1 | |

Units: `$` dollars, `%` percent, `pt` percentage points, `n` a count, `h` hours.

### The words for each row

**Win rate on best-fit accounts** (short label: Win rate)

- What: "The opportunities you already work on accounts that fit you best, and what a higher win rate on them is worth when every team works them from the same fit and the same reasons. Put in the lift you'd defend to your CFO, not one a vendor would claim."
- How: "Opportunities a year on best-fit accounts, times the average deal size, times the lift in win rate in points."

**Time back from poor-fit accounts** (short label: Time back)

- What: "The hours sellers and marketers spend on accounts that were never going to buy: research, sequences, meetings and follow-up. When the score says early which accounts don't fit, part of that time goes to accounts that do."
- How: "People who work accounts, times hours a week each spends on poor-fit accounts, times the share that comes back, times weeks worked, times what an hour costs."

**Pipeline from coordinated plays** (short label: Coordinated plays)

- What: "When marketing, sales and customer success run their plays on the same accounts, each touch lands on a buyer the others have already warmed. Price the extra pipeline at the win rate you have today, so it adds cleanly to the first row."
- How: "Pipeline a year on best-fit accounts, times the lift in pipeline you expect from coordinated plays, times today's win rate."

### Formulas

```text
win      = opps × deal × lift/100
time     = people × hours × back/100 × weeks × rate
plays    = pipe × plift/100 × win/100
total    = win + time + plays

win aside     more deals = opps × lift/100, to one decimal
              reads: "4.8 more deals won a year."
time aside    hours a year = people × hours × back/100 × weeks
              people       = hours a year ÷ 1,800, to one decimal
              reads: "2,300 hours a year, about 1.3 people's time."
plays aside   more pipeline = pipe × plift/100
              reads: "$480,000 more pipeline a year."

low           lift, back and plift × 0.5
high          lift, back and plift × 1.5, each capped at 100

program cost  entered by the person, first year, no default
payback       weeks = cost ÷ (total ÷ 52), rounded up to a whole week
              if total is 0 or less: "Doesn't pay back on these numbers."
first-year    return = (total − cost) ÷ cost, as a whole percent
```

Round dollars to whole dollars for display.

---

## ASSETS

Copy each string exactly, as one unbroken line, and put each one in the file once: a JavaScript constant that sets every `<img>` showing that image. If you can run code, pull the strings out of this file with a script instead of retyping them. If you copy by hand, check the last 12 characters against the kit before you finish.

### honeycomb-tile

AI Accelerated GTM's motif. WebP, 96 × 96. Use it in the header.

```text
data:image/webp;base64,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
```

### fresh-context-mark

The Fresh Context mark. PNG, 85 × 48, 1.76:1. Use it in the stamp.

```text
data:image/png;base64,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
```

---

## Changelog

- 1.0, 2026-10-10: first release.
