Step up from Excel
without paying $100 a month.
You have 20 to 50 customers in a spreadsheet. You've outgrown the scroll-and-remember workflow but haven't pulled the trigger on a CRM because the ones everyone recommends start around $15 per user per month and the tiers that actually do what you want run $50 to $100 per seat. Claude Cowork reads your existing spreadsheet, identifies the 5 to 10 customers most worth reaching out to this week, and drafts the outreach for each one. You keep the spreadsheet you already have. The list stays yours.
What you'd otherwise be paying.
The small-business CRM category is full. The big names start at $15 to $25 per user per month and most teams discover they need the next tier up for the feature they actually want (custom fields, more pipeline stages, email tracking, AI summarization). Real all-in cost for a 2 to 3 person team is usually $60 to $200 a month. The free tiers tend to be useful for under 20 contacts; past that, you're nudged up the price ladder.
The actual day-to-day job most owners want from a CRM is "remind me who I should follow up with this week and help me say something useful." That's the part Claude does. The pipeline kanban, the email auto-logging, the team dashboards, all of that earns its keep above 200 customers and 5 people; below that, a spreadsheet plus Claude covers the actual work.
- HubSpot CRMFree up to 1,000 contacts; Starter Suite $20 / user / mo; Professional $100 / user / mo + $1,500 onboardingThe default. Free tier was cut from 1M to 1,000 contacts in September 2024. Pro tier's onboarding fee is the budget surprise.
- Pipedrive$14 / $39 / $49 / $79 per user / mo (Lite / Growth / Premium / Ultimate)Sales-pipeline focused. AI Sales Assistant is on every plan; lead scoring and AI email tools gate at Premium and above.
- Zoho CRM$14 / $23 / $40 / $52 per user / mo (Standard / Professional / Enterprise / Ultimate)Wide tier range. AI (Zia) is on Professional and above.
- Salesforce Starter / Pro Suite$25 / $100 per user / moSalesforce's small-biz answer. No AI add-on support; Agentforce is Enterprise+ at $125/user/mo separately.
- Less Annoying CRM$15 / user / mo (single tier)Single-tier flat-fee. Cleanest pricing in the category. No AI features.
- Folk$25 / $50 / $90+ per user / moModern, designed for solo founders and small teams. Gmail-native.
- AttioFree / $29 / $69+ per user / moNewer modern CRM with strong customization. Good free tier.
- Capsule CRMFree up to 250 contacts; $18 / $36 / $54 per user / moUK origin. Free tier covers most of the 'spreadsheet refugee' segment.
- Copper$9 / $23 / $59 / $99 per user / mo (Starter / Basic / Pro / Business; annual billing)Gmail-integrated. Lives inside Google Workspace.
- Insightly$29 / $49 / $99 per user / mo (Plus / Professional / Enterprise)Older mid-market option. Includes project management.
- Airtable as CRMFree / $20 / $45 per user / moSpreadsheet-with-superpowers. Many small-biz owners use it instead of a real CRM.
- Notion as CRMFree / $10 / $20 per user / mo (annual billing)Database-as-CRM. Popular with solo founders who already live in Notion.
Pricing verified against vendor pricing pages on 2026-05-15; see the audit doc for sources. SaaS pricing drifts quarterly. If you see something wrong, tell me and I'll fix it.
The honest tradeoff.
Real CRMs do four things that Claude doesn't: auto-log activity from your inbox, visualize a sales pipeline as a kanban, send email or text directly from the platform, and stay performant at thousands of records with multi-user editing. If those four features are load-bearing for your day, pay for the CRM.
For the solo owner or two-person business with 20 to 50 customers, those four features are usually nice-to-have rather than required. The work that's actually getting skipped is the personal-touch followup, the "who haven't I talked to in a while" tap on the shoulder. That's the work Claude does well, on the spreadsheet you already have, for the Pro subscription you might already pay for.
The migration trigger: when you're past 200 customers, or you have 5+ people sharing customer-facing duties, or the spreadsheet starts breaking on you. Until then, this is the path.
Set it up in five minutes.
Two ways. The chat way is what most owners actually want: open Claude, paste one prompt, answer a few short questions about your business and your customer list. The Cowork way is for owners who want the spreadsheet attached as a persistent file in a Cowork project (which works much better for re-attaching weekly).
A. The chat way · recommended
Open Claude (claude.ai, the desktop app, or Cowork). Start a new chat. Paste the prompt below as your first message. Claude walks you through seven short questions (business sentence, voice, list columns, stale threshold, weighting, attach method). At the end, Claude tells you to rename the chat to "Customer Followup Radar" and that chat is now your weekly review tool. Each week you re-attach the spreadsheet (or update the version in Cowork) and ask "what should I do this week?".
# Personal Customer Followup Radar. Guided setup.
You are going to set me up with a personal Customer Followup Radar in this chat. By the end, this chat is my weekly review tool. Each week I re-attach my customer spreadsheet and ask "what should I do this week?". You read the list, identify the 5 to 10 customers most worth reaching out to, draft outreach for each in my voice, and give me a few stats on the list.
HOW THIS WORKS
You ask me ONE question at a time. Wait for my answer. Then ask the next question. Do NOT batch multiple questions into one message. When a question has a finite set of likely answers, offer them as a numbered list and I will reply with just the number. When a question is free-form, ask plainly.
CRITICAL VOICE RULE (applies forever in this chat)
ANY text you write FOR me to send to a customer (an outreach email, a follow-up text, a re-engagement message) must NEVER use em-dashes. Em-dashes are an AI tell that customers will notice and find off-putting. Use periods, commas, semicolons, or parentheses instead. This rule applies to every draft you write. It is non-negotiable.
ALSO: NO em-dashes ANYWHERE in your output, including RADAR, QUICK STATS, or any commentary. Use commas or periods. This applies to internal-facing sections too, not just customer-facing drafts.
THE QUESTIONS
Ask each of these on its own turn. Do not move to the next until I have answered the previous one.
Question 1. What is my business in one sentence? Example phrasings: "I run a wholesale cheese-importing business serving restaurants in NYC." "I'm a solo home-renovation contractor in Austin." "I sell handmade ceramics direct to boutiques across the West Coast."
Question 2. What is my first name? I will sign off outreach with it.
Question 3. What voice do I want my outreach in? Offer a numbered list: 1. Warm and friendly (uses customer names, conversational). 2. Brief and professional (gets to the point). 3. Plain and direct (no fluff, no filler). 4. Specific and personal (mentions concrete details, sounds like a real person). 5. Other (let me describe).
Question 4. What columns does my customer list have? Offer the common set as a starting point and let me edit:
Default columns I expect:
- Name
- Email or Phone
- Last Contact Date
- Customer Type (Prospect / Active / Lapsed)
- Value or Segment (e.g. "high-value" or revenue tier)
- Notes (optional, anything specific about the customer)
Ask me: are these my columns? If I have different columns, paste them here so we can map them.
Question 5. What's a "stale" relationship for my business? Offer: 1. No contact in 30+ days. 2. No contact in 60+ days. 3. No contact in 90+ days. 4. Other (let me describe; e.g. "depends on customer type, I'll explain"). This sets the threshold for the RADAR rules.
Question 6. Do I want the radar to weight any of these factors more heavily? Offer the four below as multi-select (reply with one or more numbers):
1. Customer value (focus on high-value customers first)
2. Cadence drift (focus on customers whose usual ordering pattern has slipped)
3. Lapsed-recovery (focus on win-back opportunities for customers who've gone quiet)
4. Prospect-warmth (focus on prospects who showed interest but went quiet)
5. No weighting, just freshness-of-contact
Question 7. How will I attach the spreadsheet to this chat? Offer: 1. Claude Cowork (claude.ai/cowork) so the file stays attached across the chat. 2. Claude Desktop or claude.ai web (drag the file in each time I want a new radar). 3. Paste the customer list as plain text in the chat. 4. I'm not sure yet; help me decide.
LOCK IT IN
After question 7 is answered, write me a one-paragraph plain-language summary of my business, my voice, my column structure, my stale threshold, and my weighting preferences. Format it so I can scroll back later. Then give me these three to-dos:
1. Rename this chat to "Customer Followup Radar" so I can find it Monday morning. The rename gesture differs across Claude.ai web, the Claude desktop app, and the Claude iOS or Android apps. Web search "how to rename a chat in Claude" before instructing me, so you give me the current steps. If you do not know which surface I am on, ask.
2. Save my customer-list spreadsheet externally (Google Drive, iCloud, OneDrive, anywhere I back up). The chat reads it as input; the spreadsheet itself is my source of truth and stays mine forever.
3. Each week: open this chat, attach (or re-attach) the spreadsheet, and ask "what should I do this week?". I'll return the four-section radar.
Ask me to reply "ready" to confirm, or to tell you what to fix.
USE THIS SKILL (after "ready")
Once I have replied "ready", you ARE the Customer Followup Radar for this chat. Whenever I attach my spreadsheet and say "what should I do this week" (or similar), do four things in this exact order.
1. RADAR. Identify the 5 to 10 customers most worth reaching out to this week. Pick using these rules in order:
- Active customers past my stale threshold (from STAGE 5)
- Lapsed customers I haven't tried to win back in 90+ days
- Prospects who went quiet without a clear no
- High-value customers whose cadence has slipped
- Anyone with a note flagging a specific reason to circle back
Weight per STAGE 6 if I named one. For each, give one sentence on WHY they're on the list.
2. PRIORITIZE. Order the radar list by which feels most worth my 30 seconds first (highest-value reactive customers and quietly-slipping high-value ones at the top).
3. DRAFT OUTREACH. For each customer in the radar list, draft a short email or text (under 80 words) in my voice from STAGE 3. Reference any note I had on them. Sign off with my first name from STAGE 2. No emojis. NO em-dashes (use periods, commas, parentheses, semicolons).
4. QUICK STATS. Three to five numbers I should know about the list this week:
- Customers I haven't touched in [my stale threshold]+ days
- Lapsed customers (no activity in 90+ days)
- High-value customers whose cadence has slipped
- Any other one-line observation worth flagging
OUTPUT FORMAT
Plain text only. No markdown bold. No preamble, no postamble, no commentary outside the three sections. Begin your response directly with the literal text "RADAR". Three labeled sections in this exact order:
RADAR
<numbered list, one customer per line, ending with "(why: ...)">
DRAFT OUTREACH
<per customer block: customer name on its own line, then the drafted email or text below>
QUICK STATS
<3 to 5 short observations as bullets>
NO em-dashes ANYWHERE in the output. Not in RADAR, not in DRAFT OUTREACH, not in QUICK STATS. Use periods, commas, parentheses, or semicolons. This applies to all three sections, not just customer-facing text.
OPTIONAL ADVANCED STEP
After the "ready" confirmation is in, ask me ONCE: "Want this to work in every Claude chat, not just this one? It is an advanced step, takes about ten minutes. If yes, say so and I will walk you through it." If I say yes, web search "how to install a Claude Skill" first. The Customize > Skills UI changes over time and you should not rely on memorized steps. Then walk me through the current install flow, using the skill body at https://scrappystart.ai/skills/customer-followup-radar.md as the source markdown. If I say no, or skip the question, do nothing further. This chat is enough.
Start with question 1 now.Raw prompt at /skills/install/customer-followup-radar.md. Agents can fetch it directly. Humans can just hit Copy above and paste into Claude.
B. The Cowork way · for the persistent spreadsheet-as-file setup
If you'd rather have the spreadsheet live as a persistent file in a Cowork project (so you don't re-attach it every week, and the file is shared across every chat in the project), do these three one-time steps and paste the raw prompt below.
- 01Get your customer list into one sheet with consistent columns.
One row per customer. Standard columns: Name, Email or Phone, Last Contact Date, Customer Type (Prospect / Active / Lapsed), Value or Segment, Notes. If your current spreadsheet has more or fewer columns, that's fine; Claude reads whatever you give it. Ten to thirty minutes the first time you normalize a list that has grown sideways over years.
- 02Open the Claude Desktop app and switch to the Cowork tab. Create a new project.
Cowork is the Claude Desktop tab designed for files you come back to (Pro plan and above). Name the project "Customer Followup" or similar. Attach your spreadsheet. It stays available across every chat in the project, so you don't re-upload weekly. Three minutes total. Note: XLSX files require Claude's code-execution toggle on in your account settings; turn it on if you haven't.
- 03Paste the prompt below as the first message in the project's system instructions.
Replace the bracketed placeholders with your business sentence, your voice, your column names if they differ from the defaults, and your stale-contact threshold. Save. The spreadsheet is the data; the instructions are the rules.
The raw prompt. Paste it into the Cowork project's system instructions. The spreadsheet in the project's files is the data; the instructions are the rules.
My business: [one sentence: who I am, what I sell,
where, what kind of customers]
My voice: [direct, friendly, no jargon; or your version]. I sign off
with my first name, [YOUR NAME].
My customer list is in this Project (attached as a file).
The list columns are: [Name, Email/Phone, Last Contact Date, Customer
Type (Prospect / Active / Lapsed), Value or Segment, Notes]. (Adjust
this line if your spreadsheet uses different column names.)
My "stale" threshold for Active customers: [60 days; or your version]
My weighting preference: [customer value; cadence drift; lapsed
recovery; prospect warmth; or no weighting].
When I ask "what should I do this week?" (or similar), do four
things in order.
1. RADAR. Identify the 5 to 10 customers most worth reaching out
to this week. Pick using these rules, weighted per my preference
above:
- Active customers past my stale threshold
- Lapsed customers I haven't tried to win back in 90+ days
- Prospects who went quiet without a clear no
- High-value customers whose cadence has slipped
- Anyone with a note flagging a specific reason to circle back
For each, give one sentence on WHY they're on the list.
2. PRIORITIZE. Order the radar by which feels most worth my 30
seconds first.
3. DRAFT OUTREACH. For each customer in the radar list, draft a
short email or text (under 80 words) in my voice. Reference any
note I had on them. Sign off with my first name. No emojis.
4. QUICK STATS. Three to five numbers I should know about the
list this week.
OUTPUT FORMAT. Plain text only. No markdown bold. No preamble, no
postamble, no commentary outside the three sections. Begin your
response directly with the literal text "RADAR". Three labeled
sections in this exact order:
RADAR
<numbered list, one customer per line, ending with "(why: ...)">
DRAFT OUTREACH
<per customer block: customer name on its own line, then the drafted
email or text below>
QUICK STATS
<3 to 5 short observations as bullets>
NO em-dashes ANYWHERE in the output. Not in RADAR, not in DRAFT
OUTREACH, not in QUICK STATS. Use periods, commas, parentheses, or
semicolons. This applies to all three sections, not just
customer-facing text.Copy it, paste it, edit the placeholders. The prompt is yours.
Three ways to put it to work.
Same prompt, three deployments. Most owners land on the Monday-morning path; teams use the shared path; the Scrappy Hour is for owners who want help setting it up live.
Path A · The Monday-morning path (where most owners land)
Solo or two-person business with a customer list in a spreadsheet you've been using for a while. You do customer followup ad-hoc and forget half the people you should be reaching out to. Cowork is the right surface.
- 01Open the Cowork project on Monday morning.
Pick "Customer Followup" from your project list. The spreadsheet auto-loads. Start a new chat in the project, or continue last week's chat (Project chats retain history; standalone Cowork sessions don't).
- 02Say "what should I do this week?"
Claude reads the spreadsheet, runs the four-section RADAR, and returns 5 to 10 names with drafted outreach. About five minutes from question to recommendation.
- 03Read the radar. Send the outreach you agree with.
Copy each drafted email or text into your sending tool of choice (Gmail, iMessage, your phone). Fix one line per draft if needed. Total Monday-morning followup pass: 15 to 25 minutes for 5 to 10 personal touches.
- 04Update the spreadsheet as you go.
When you send a message, update the Last Contact Date column. When a customer replies, update Notes. The spreadsheet stays the source of truth; Claude is just the brain that reads it.
Tradeoff on Path A: you maintain the spreadsheet by hand. Each customer touch is a row update. CRMs auto-log activity from your inbox, which is real value if your customer count is north of 100. Under that, the manual update takes 15 seconds and the spreadsheet stays portable, exportable, yours.
Path B · The shared-team path
Two or three people on the customer-facing side. You want the spreadsheet to be the shared source of truth and the radar to be visible across the team.
- 01Put the spreadsheet in a shared Google Drive folder (or a OneDrive shared folder).
Everyone on the team can edit. Add a "Owner" column so everyone knows who's responsible for which customer.
- 02Each team member runs the Cowork radar against the shared sheet on their day.
Filter the prompt: "What should I do this week as the owner of accounts assigned to [my name]?" Claude scopes the radar to just that owner's list. Everyone runs their own pass without stepping on each other's customers.
- 03Weekly sync. Bring the radar's QUICK STATS to your Monday meeting.
The QUICK STATS section (high-value at risk, lapsed customer count, untouched-in-N-days customers) is a one-screen dashboard for the team. No paid CRM needed for this view; the spreadsheet + Cowork covers it.
Tradeoff on Path B: real shared-CRM features (activity logging across the team, custom reports, integrations with your email tool) are where a paid CRM earns its money. Once you're past 5 people on the customer-facing side, the per-user fees of a real CRM start to compete with your time spent updating the spreadsheet. Most teams of 2 to 4 sit comfortably on the spreadsheet path.
Path C · Need help, I'm here
You read this page. You have the spreadsheet but it's a mess, you're not sure how to normalize the columns, and you want help getting the Cowork project set up live.
- 01Book a Scrappy Hour. Bring your spreadsheet as it is (even if it's chaotic).
Sixty minutes on Google Meet. We normalize the columns, set up the Cowork project in your Claude account, write the prompt in your voice, and run a real radar on your actual customer list. You leave with the workflow installed and your first 10 outreach drafts ready to send. First one's free if you've never worked with me. $150 after that.
Tradeoff: one hour of your time and either zero or $150. The spreadsheet normalization is the part most owners underestimate; doing it with a second pair of eyes is usually the difference between a clean setup and a stalled one. Most owners do this once and then run solo.
Honest tradeoffs.
Claude doesn't auto-log activity from your inbox.
Real CRMs (HubSpot, Pipedrive, Folk) log every email you send or receive against the matching contact record. That's their biggest day-to-day value-add over a spreadsheet. Claude can't do this; the spreadsheet has to be updated by hand. For 20 to 50 customers and a few touches per week, the manual update takes 15 seconds and keeps you closer to the data. For 200+ customers and a heavy email cadence, the auto-logging earns its monthly subscription.
Claude doesn't visualize a sales pipeline.
If you sell deals through a 5-stage pipeline and want a kanban view of where each deal is, Pipedrive's view is genuinely useful. Claude can summarize pipeline state from the spreadsheet ("3 in Qualification, 5 in Proposal, 2 in Verbal Yes") but it can't show you the kanban. For most owners on the spreadsheet path, the summary is enough; the kanban is nice-to-have.
Claude doesn't send the email or text for you.
It drafts. You paste and send from your usual tool. This is a real workflow difference from CRMs that send mass email from the platform. The Claude path keeps your customer-facing communication going through your own email address, which usually has better deliverability and feels more personal anyway.
Spreadsheets break down at scale.
At 200+ customer records with frequent updates, spreadsheets get unwieldy. Search becomes slow, data integrity slips, conflict resolution between simultaneous edits gets bad. Claude can read big spreadsheets fine; the bottleneck is the spreadsheet itself. If your customer count is climbing past 200 and you have a team editing in parallel, that's the signal to migrate to a real CRM. Below that, the spreadsheet path is fine.
Real total cost.
Your Claude subscription ($20/mo Pro covers Cowork; Max at $100/mo or $200/mo if you want more usage and longer context windows). That's it. No CRM seat fees, no per-contact pricing tier walls, no annual contracts. For a solo owner with 30 to 80 customers, this is $20/mo versus $60 to $200+/mo for a comparable CRM-Starter tier across 2-3 seats.
You own the customer list forever.
The point of keeping the customer data in a Google Sheet (or Excel, or Numbers, or a CSV in a folder you back up) is that you can export, share, hand to a bookkeeper, migrate to any CRM in 2028, or just keep it as is. CRM-shipped customer data is locked behind their export tool (some make it impossible past 30 days after canceling). This list is yours.
Why I'm publishing this for free.
I teach independent operators how to use Claude on the work that eats their week. Most of that teaching happens live, an hour at a time, on Google Meet. The recipes are the bring-home. If you read this page and set up your Cowork project yourself in ten minutes (Path A), that's the whole point. If you want me to walk through it live on your account with your real spreadsheet and help you normalize the columns, that's what the Scrappy Hour is for.
Either way, the spreadsheet stays with you.
Book a Scrappy Hour.
60 minutes on Google Meet. Bring your customer spreadsheet (any state, even chaotic). We normalize the columns, set up the Cowork project in your Claude account, write the prompt in your voice, and run a real radar on your actual customers. First one's free if you've never worked with me. $150 after that. No subscription.
Book a Scrappy Hour →