CSM Interview · Take-Home War Room

The Save.
Not the downgrade.

Harshleen — Bob wants out. Your job isn't to process a cancellation, it's to turn a churn email into a 90-day growth plan using his own data. This kit is your scaffold: the thesis, the formulas, the email, the 6-slide blueprint, and the delivery notes. The thinking stays yours.

Account: Bob's Burgers · 5 locations Package: Pro · $205 / store / mo Annual contract value: $12,300 Deadline: 48 hours · PDF + CSVs
⬇ Download the PowerPoint starter deck

What they are actually testing

A cancellation email is a gift. It's a chance to show you lead with value, not concessions.

They told you the constraint on purpose: you cannot approve a downgrade or cancellation. That removes the easy exit and forces the real skill — can you de-escalate, find the story in the data, and re-anchor the relationship on outcomes? Here's the rubric hiding inside the prompt.

The 5 things they're scoring read between the lines

  • De-escalation under constraintDo you reflexively concede, or hold the line warmly and earn the conversation?
  • Data literacyCan you turn two CSVs into one clear narrative — not a number dump?
  • Value articulationCan you reframe "cost" as ROI, owned-channel value, and assets he'd lose?
  • Executive communicationIs the email empathetic, specific, and commercially confident in equal measure?
  • OwnershipOne next action, one named owner, a date. No vague "we'll circle back."

Decode Bob's message the real problem

"The product seems fine, but I'm just not seeing results." In CSM, "not seeing results" almost never means the product failed. It decomposes into three diagnosable causes — and your CSVs will tell you which:

  • Adoption gapHe pays for Pro (CRM email, loyalty, custom app) but isn't using the revenue features. Check campaign volume & cadence.
  • Attribution blindnessSales may be flat-to-up vs his old provider, but he has no clean before/after, so it "feels" like nothing.
  • Wrong baseline / expectationsJoined January — owned-channel ordering ramps. ~5 months is early, and no one set 90-day targets with him.

Your one-sentence thesis (build everything around this): Bob doesn't have a product problem — he has an adoption-and-visibility problem, and his own data proves the Pro package is already creating value he can't see. Before any contract change, we activate the unused features and make the ROI visible on a 90-day plan.

The submission, mapped to 6 slides

Max 6 slides total. Part 1 is 3–4, Part 2 is 1–2. Don't waste a slide on a title page — make every one carry weight.

Part 1 · Slide 1

Performance summary + your diagnosis

Part 1 · Slide 2

Value vs cost (the money slide)

Part 1 · Slide 3

90-day strategy + the single next action & owner

Part 1 · Slide 4

The exact email (verbatim)

Part 2 · Slide 5

Top 3 book-of-business insights

Part 2 · Slide 6

How they shape your prioritization

Tip: the CSV analysis lives in your .csv deliverables — slides show the conclusion, the CSVs show the work.

Turn two CSVs into one story

Don't compute everything. Compute the five numbers that change the decision.

You have Bob's Burger Sales Data.csv and Bob's Burger CRM Campaign Data.csv. Below are the exact metrics that build the "you're getting value you can't see" case, plus copy-paste Google Sheets / Excel formulas. Adapt the column letters to your file (use the inspector at the bottom to map them fast).

The 5 numbers that win the argument

  • Sales trend since January — is online ordering growing month over month?If yes: "your channel is compounding." If flat: it's the adoption gap, which you can fix.
  • Owned-channel value (commission avoided) — first-party orders carry no marketplace commission.Often the single number bigger than the fee. This is the headline of your value-vs-cost slide.
  • CRM engagement vs potential — how many campaigns has he actually sent?The likely smoking gun: low/zero campaigns = low results. That's not failure, that's runway.
  • Repeat / loyalty rate — is the loyalty program creating return visits?Repeat customers are the asset he'd forfeit by leaving.
  • Per-location spread — are some of his 5 stores thriving and some dormant?Variance proves it's executional, not the product. Lift the laggards to the leaders.

Copy-paste formulas Sheets / Excel

Month-over-month growth %Is the channel compounding?
=(ThisMonth - LastMonth) / LastMonth
Total & per-location salesSUMIFS by store + month
=SUMIFS(SalesCol, LocationCol, "Store 1", MonthCol, "Feb")
Average Order Value (AOV)Revenue ÷ order count
=SUM(SalesCol) / COUNTA(OrderIDCol)
Repeat-customer rateLoyalty working?
=COUNTIF(CustomerIDCol, ">1 order") / COUNTA(UniqueCustomers) // or with a helper: returning customers ÷ total customers
Commission avoided (owned-channel value)Flag your % assumption on the slide
=OwnedChannelSales * 0.20 // 15–30% typical marketplace take rate
Email funnel: open / click / conversionFrom the CRM CSV
OpenRate =Opens / Delivered ClickRate =Clicks / Opens ConvRate =OrdersFromCampaign / Clicks
Campaign ROIRevenue per dollar of effort
=(AttributedRevenue - CampaignCost) / CampaignCost
Value-vs-cost ratio (the headline)Return per $1 of package spend
=MonthlyOnlineSales / (5 * 205) // 5 stores × $205 = $1,025 / mo
Project the next 90 daysTrendline forecast
=FORECAST.LINEAR(NextMonthX, KnownSalesYs, KnownMonthXs) // quick visual: =SPARKLINE(MonthlySalesRange, {"charttype","line"})
The reframe that wins: stack his real value in one column — online sales + commission avoided + loyalty repeat revenue + the customer list he now owns — and put $1,025/mo next to it. If the value column clears the cost (it usually does), the cancellation conversation is over. If it doesn't yet, that's exactly what the 90-day plan closes.

Live value-vs-cost calculator play with it

Drop in Bob's real numbers to pressure-test your slide before you build it.

8.8×
return per $1 of package spend
Monthly package cost$1,025
Commission avoided / mo$1,800
Commission alone vs feecovers 176% of cost
Net value after package / mo$7,975

This is a directional model to sharpen your instinct — your CSV is the source of truth. Always state assumptions (commission %, attribution) on the slide.

CSV inspector map your columns fast

Drop either CSV here. It runs entirely in your browser (nothing uploaded) and shows every column, type, totals, averages and a quick trend — so you know exactly which column letters to feed the formulas above.

Drop a .csv here or click to choose 100% local · nothing leaves this page

The exact reply you'd send

Warm. Specific. Confident. You're not negotiating — you're inviting him back to his own numbers.

This is a ready template. Swap the [bracketed] bits for the real findings from your CSVs once you've run the formulas. The annotations below explain why each move works — be ready to defend them in the interview.

Why each move works be ready to explain

01
Thank him for the honesty first. Disarms the tension and signals you're not defensive. Never argue with the feeling.
02
"Before we talk about any change…" Holds the line on no-downgrade without saying "no." You redirect, you don't refuse.
03
Lead with his own data. Specificity = credibility. The moment you cite a real number, you stop being a vendor and become his analyst.
04
Name the unused features. Reframes "not working" as "not switched on yet" — a problem you can visibly solve, which restores hope.
05
One concrete next step + a date. A 30-minute session is a low-commitment yes. You own the prep; he just shows up.
06
Confident close, no begging. "I'll bring the data and a plan" projects ownership. You're leading the save, not pleading for it.
What to avoid: apologizing for the price, explaining contract/cancellation mechanics, over-promising specific revenue, or sending a wall of charts. The email's only job is to win the 30-minute meeting.

6 slides, max — every one earns its place

Slides show the conclusion. Your CSVs show the work.

A slide-by-slide blueprint with the headline, the supporting points, and the suggested visual for each. Use a clean, branded template; one idea per slide; let charts do the talking.

1
Part 1 · Performance + Diagnosis

"You're growing — you just can't see it yet."

Headline the trend and per-location spread. State your diagnosis in one line: adoption + visibility gap, not a product failure.
Line chart: monthly online sales since Jan · small bar chart: sales by store
2
Part 1 · Value vs Cost

"The math already works."

Stack the value column — online sales + commission avoided + loyalty repeat revenue + owned customer list — beside $1,025/mo. Show the return-per-dollar ratio.
Side-by-side bar: total value vs monthly cost · callout number: "X× ROI"
3
Part 1 · Strategy + Next Action

"Here's the 90-day plan to make it undeniable."

Activate unused Pro features → set 30/60/90 targets → biweekly check-ins → QBR at day 90. Then the single next action: book a Value & Growth working session within 72 hours — owner: you (CSM).
90-day timeline · RACI line for the one action
4
Part 1 · The Email

The exact reply to Bob (verbatim)

Drop the email from tab 03 in clean, readable type. This is a deliverable they explicitly asked for — present it as a finished artifact, not a screenshot.
Formatted email block · signature
5
Part 2 · Book of Business

Top 3 insights from your 100 merchants

e.g. revenue concentration (few Mx = most revenue) · adoption predicts retention (low-usage Mx are the churn risk) · a renewal/at-risk cluster needing proactive plays. Back each with one number.
2×2 value-vs-risk matrix · Pareto bar of revenue by Mx
6
Part 2 · Prioritization

"How I'd spend my week because of it."

Translate insights into action: protect high-value at-risk (Bob lives here), grow healthy high-value via expansion, scale/automate the long tail. Tie it back to a repeatable playbook.
Effort-allocation donut · the playbook in 3 lanes
Design discipline: one headline per slide written as a claim (not a label), max ~5 lines of support, a single chart, consistent color. If a slide has two ideas, it's two slides — or one of them is cut.

Part 2 · the CSM Book of Business CSV

Insight isn't "I have 100 merchants." It's "so here's where my hours go."

You won't know the columns until you open it, so here's the framework that works on any book of business — plus the formulas to build it and the matrix to present it.

Step 1 — Build a simple health score

Normalize a few signals to 0–1 and weight them. Tune weights to whatever columns the CSV actually gives you.

Weighted health scorehigher = healthier
Health = 0.30*Revenue_n + 0.25*Adoption_n + 0.20*Sentiment_n + 0.15*Tenure_n + 0.10*GrowthTrend_n // _n = normalized 0–1, e.g. =(x-MIN)/(MAX-MIN)
Revenue concentration (Pareto)find the vital few
=SUMIF(RevenueCol, ">="&LARGE(RevenueCol, 20)) / SUM(RevenueCol) // what % of revenue do your top 20 accounts hold?
Segment countshow big is each quadrant?
=COUNTIFS(RevenueCol, ">="&MedRev, HealthCol, "<"&MedHealth) // high value + low health = your Protect list

Step 2 — Drop every merchant into the matrix

Value (revenue / package tier) on one axis, Risk (low health, near renewal, low adoption) on the other. This single picture is your prioritization slide.

High value · High risk
🛡 Protect
Proactive saves. White-glove. Bob's Burgers lives here. Your highest-leverage hours.
High value · Healthy
🌱 Grow
Expansion & advocacy. Upsell, references, case studies, multi-year renewals.
Low value · High risk
⚖ Triage
Scaled outreach. Don't over-invest; right-size effort to value, accept some churn.
Low value · Healthy
🔁 Automate
Tech-touch & self-serve. Nurture cheaply; watch for ones quietly becoming Grow.

Step 3 — The 3 insights to actually surface

Open the CSV, then look for these patterns. Lead each slide bullet with the number, not the adjective.

  • Concentration of value"My top [N] merchants drive [X]% of revenue" → they get named, scheduled, protected.
  • Adoption predicts retention"Low-adoption accounts show [the churn signals]" → proactive activation plays before they ever email like Bob.
  • A timing cluster"[N] accounts renew in the next quarter / are newly onboarded and unactivated" → a focused, calendar-driven motion.

Whatever the exact columns, the move is the same: from "100 accounts" to "here are the 8 I touch this week and why."

The close for Part 2: "My book isn't 100 equal accounts — it's a portfolio. I spend my time where value and risk are both high, I run scaled plays everywhere else, and Bob's Burgers is a textbook Protect account, which is exactly why I'd never just process his downgrade."

Stick the landing

They're not buying the deck. They're buying how you think out loud.

Use this kit the way the prompt intends. They said AI is fine for outlining, clarity, and proofreading — and that the core problem-solving, decisions, and conclusions must be yours. So: run your own formulas on the real CSVs, draw your own conclusions, write the email in your voice. Let this kit be the scaffold, not the substance. You'll be asked to explain your approach live — make sure every number on a slide is one you computed and can defend.

How to present it

  • Open with the thesis, not the data"Bob doesn't have a product problem, he has a visibility problem — and his data proves it." Then earn it.
  • Every number ties to a decisionIf a chart doesn't change what you'd do, cut it.
  • Empathy + commercial backboneShow you can be warm to Bob and protect $12,300 of ARR at the same time.
  • End on ownershipOne action, your name on it, a date. Leave no doubt who drives the save.

Likely interview questions

"What if Bob still wants to cancel after the meeting?"
"Then I've earned the right to escalate options — a tailored success plan, a temporary commercial gesture if policy allows, or a graceful path. But I don't open with concession; I open with value, because most 'cancel' emails are really 'help me' emails."
"What if the data shows he's genuinely underperforming?"
"Then the gap is my strongest asset — it's the exact runway the 90-day plan closes. I'd benchmark him against similar merchants, isolate the unused levers, and set targets we both sign. Underperformance with idle features is the most fixable problem there is."
"How did you pick your assumptions?"
"I flagged every assumption on the slide — commission rate, attribution window — and used conservative figures so the case holds even at the low end. I'd refine them with his actuals in the working session."
"Why is the next action yours and not support's?"
"Because this is a relationship moment, not a ticket. I own the save; I'll pull in enablement to activate features, but Bob needs to hear the plan from the person accountable for his success."

Your AI-use statement have this ready

If they ask how you used AI, be specific and confident — it reads as maturity, not weakness:

"I used an AI tool to pressure-test my structure and tighten wording — outlining the slide flow and proofreading my email. The analysis, the diagnosis, the strategy and every conclusion are mine: I ran the formulas on the CSVs, decided the thesis, and chose the next action. I can walk through any number on these slides without it."

Pre-submit checklist

  • ≤ 6 slides, exported as one PDFPart 1: 3–4 · Part 2: 1–2
  • .csv files attached with your analysisThe work behind the conclusions — clean tabs, labeled columns
  • The verbatim email is on a slideIt's an explicitly requested deliverable
  • One next action + named owner + dateDon't let this be vague
  • Every assumption labeledCommission %, attribution, normalization
  • Submitted via the link, inside 48 hoursEarly is a signal too
You've already managed a $3M book and saved real renewals. This is that, on a smaller stage with a burger logo. Go show them the operator you already are. 🍔

Walk in fluent · verified from DoorDash's own docs, June 2026

Know exactly what Bob bought — and the commercial model behind it.

The role is CSM for Online Ordering — DoorDash's first-party SaaS product, which is a different animal from the Marketplace app. Knowing the difference is how you sound like an insider, not a candidate. Here's the verified landscape.

The product you'd own first-party

Online Ordering / Storefront lets a restaurant take orders on its own website and branded app — commission-free, paid as a SaaS subscription instead of a per-order marketplace cut. Bob's Pro package ($205/store) bundles the full stack:

  • Website online ordering + custom mobile appHis own branded channel — he owns the customer, not the platform
  • CRM email marketingDirect-to-customer campaigns — the unused lever in Bob's case
  • Custom loyalty programRepeat-visit engine and a customer list he keeps

This is why your value-vs-cost case is strong: first-party orders dodge the 15–30% marketplace commission entirely. The package is the owned channel.

The Marketplace tiers context

The other way restaurants sell on DoorDash — the app itself, priced on commission. Good to know so you can speak to the whole portfolio:

  • Basic — 15% delivery commissionSmallest delivery radius, 7-day trial
  • Plus — 25% delivery commissionLarger radius, DashPass, 30-day trial
  • Premier — 30% delivery commissionBiggest radius + Automatic Ads + $200 photo credit + Growth Guarantee (refunds commission in any month with ≤20 orders)

All tiers: flat 6% pickup commission, $0 activation, free trial, and a free professional menu photoshoot.

What a merchant actually receives — physical & digital

Digital · 3 surfaces
  • Merchant PortalDesktop menu editor — items, photos, modifiers, hours
  • Business Manager appiOS/Android — live order management, refunds, subs
  • Order Manager appThe tablet app — pop-up on each new order
Hardware
  • 10" Android tabletFree during trial, then $6/week in the US
  • Ships in ~7 daysAfter the confirmation email; opt-in, not bundled
  • It's the order hubFor merchants without a POS integration
Marketing materials
  • The tablet is the only confirmed shipped itemper DoorDash's own docs
  • Printed kit is order-on-requestDoorDash runs a "marketing materials" ordering program; exact contents (decals, table tents, QR cards) vary by region — verify, don't assume auto-ship

Sourced from DoorDash's pricing page, learning center, and help center via a 102-agent fact-check (all claims verified against primary docs). Figures are US, 2025–2026, and DoorDash can change them.

How to use this live: when you reframe Bob's value, name the mechanism — "your Storefront orders are first-party, so unlike the DoorDash app they carry no 15–30% commission; that owned channel plus your CRM list is the asset you'd be walking away from." That's the sentence that makes an interviewer lean in.

~$105K in Manhattan · what it actually means

$105K sounds big. Manhattan quietly takes ~30% in tax and rents most of the rest.

Walk into the comp conversation knowing the real number. These are directional 2026 estimates for a single filer — confirm your exact figure with a paycheck calculator (e.g. SmartAsset). But the shape is right, and it changes how you negotiate.

Live take-home & budget calculator NYC, single filer

Federal income tax
NY State tax
NYC city tax
FICA (SS + Medicare)
Total tax
$6,114
estimated take-home per month
Net per year$73,367
Effective tax rate30.1%
Comfortable rent (≤30% net)$1,834
50 / 30 / 20 monthly$3,057 · $1,834 · $1,223

"Comfortable rent" is the classic 30%-of-take-home guideline. Note how far below a solo Manhattan apartment that sits — which is the whole point.

Three realistic ways to live on it

Solo · Manhattan

Studio ~$3,300–4,200/mo

Eats 55–70% of take-home. Doable only if you forgo most saving. The "I made it" flex that quietly breaks the budget.

Share · prime

Room in 2BR ~$1,900–2,600/person

Manhattan or brownstone Brooklyn with a roommate. ~31–43% of take-home. The realistic sweet spot for most.

Solo · commute

Studio/1BR ~$1,800–2,600/mo

Astoria, Washington Heights, Bushwick, Jersey City. Your own place, ~30–43%, in exchange for a train ride.

The reality check purchasing power

Manhattan's cost of living runs roughly 2.2–2.4× the U.S. average, and rent closer to 3–4×. In lifestyle terms, $105K here lands near a $45–55K salary in an average U.S. metro — directionally, not exactly.

~30%
to taxes (fed+state+city+FICA)
~$6.1K
take-home / month
~40%+
of that to rent, realistically

None of this means turn it down — NYC pays a premium for a reason. It means negotiate from facts, not the headline number.

Negotiation playbook

  • Is $105K base or OTE?CSM roles often have a variable component. On-target earnings ≠ guaranteed base. Clarify first — it's the whole ballgame.
  • Ask for the band"What's the range budgeted for this level?" Anchors you to their ceiling, not their opener.
  • Bring compsLevels.fyi, Glassdoor, Built In NYC, RepVue (CS/sales). NYC mid-level CSM base commonly ~$90–125K + variable.
  • Negotiate the whole packageSigning bonus, hybrid/remote days (a real $ saving on commute), transit & wellness stipends, extra PTO, a 6-month review, equity/RSUs.
  • Use your leverageYou manage a $3M book today. You're not a junior — price accordingly.
What to actually say: "I'm genuinely excited about the role and the product. Based on my experience managing a $3M enterprise book and the NYC market for this level, I was targeting a base in the [$115–125K] range. Can we get there — and could we also look at [sign-on / equity / hybrid days] to round it out?" Calm, specific, collaborative. Name a number; let them counter.

Estimates only — not tax or financial advice. Brackets are 2025–2026 figures for a single filer taking the standard deduction; your real number depends on benefits, 401(k), filing status and more.

Your deliverable scaffolds

Polished starting points — then make every number and conclusion yours.

Read this first. The PowerPoint is a designed template, and the CSVs are worksheets — structure, prompts and formulas, not answers. The assignment requires your own analysis and conclusions, and you should be ready to explain your approach. Fill in every [bracket] with figures you computed from the real CSVs, and rewrite the copy in your voice.

① The PowerPoint deck 6 slides · 16:9 · editable

A branded, ready-to-edit starter matching the 6-slide blueprint: Situation & Diagnosis · Value vs Cost · Strategy & Next Action · The Email · Book of Business · Prioritization. Every slide has placeholder charts and speaker-note coaching built in.

⬇ Editable .pptx ⬇ Preview as PDF

Open the .pptx in PowerPoint or Google Slides → replace sample charts with your CSV data → export to PDF for submission. Check the notes pane under each slide. (The PDF is just a quick look at how it renders.)

② Part 1 — Sales analysis worksheet .csv

Every metric that builds the value case, with the exact Sheets/Excel formula and a blank column for your result — sales trend, AOV, commission avoided, per-store spread, CRM funnel, ROI.

⬇ Part1_Sales-Analysis-Worksheet.csv

③ Part 2 — Book of Business worksheet .csv

The health-score formula, the 2×2 segmentation counts, and the three-insight scaffold — drop in your CSM Book of Business numbers and the prioritization writes itself.

⬇ Part2_Book-of-Business-Worksheet.csv

④ The whole toolkit as a PDF read it offline

Every tab of this site — the brief, formulas, email, blueprint, book of business, product briefing, NYC pay — in one scrollable PDF, in case you'd rather read it on your phone than click around.

⬇ HK-CSM-Toolkit-FULL.pdf
Submission checklist: export the finished deck to one PDF (≤6 slides) · attach your .csv analysis · make sure the verbatim email is on a slide · one next action + named owner + date · every assumption labeled · submit via the link inside 48 hours.