CRM or Spreadsheet: When You Outgrow Excel

In short: a spreadsheet is a fine tool while you close under 10 deals a month with one seller in the room. Past that, every next deal costs more to serve: leads leak between chats, “who sent the quote?” becomes an email archaeology project, and reports are assembled by hand. A custom CRM pays back not through a pretty dashboard, but by deleting those costs. We quantify them before starting — the CRM consultation is free.

The threshold is not deal count, it is hand count

A seller who runs his own sheet knows it by heart: what sits where, why row 47 is empty, who to call on Friday. The file works because the only reader and the only writer are the same person.

The break comes when the eleventh deal arrives — no, when a second person starts typing in the same file. The first version conflict (“I deleted that row yesterday, you opened the wrong copy”) is not a disaster yet. It is a signal: the sheet stopped being the single source of truth. From that moment every number carries a date and an author’s name, and the owner’s morning goes into working out which of the three files is the real one.

The second threshold is channels. One row can start in Instagram, continue in Viber, close on a call and get paid through email. A sheet records only the last step, so you see “the deal happened” and never see the four that died between steps.

Five signs you already outgrew Excel

1. Leads leak between channels. An Instagram enquiry never reached the sheet — you will never hear about it. In the Ovisko case it looked like three services that could not see each other.
2. “Where did we stop?” costs 15 minutes. If the communication history lives in a manager’s phone instead of a system, the deal stalls when they take vacation.
3. A report = a day of work. Leadership sees last month’s numbers only when the bookkeeper finishes merging files. A CRM shows margin in real time.
4. Manual routine. Copy data from an email into a request, generate an invoice, remind about payment — three operations per deal. Two hundred deals a month at four minutes each is 13 hours, five working days nobody notices because they are smeared thin across the whole week.
5. Processes cannot grow. A new product line, a new department, a 1C integration — and the spreadsheet breaks, because it does not scale.

Where those figures come from. Speed-to-lead is the InsideSales and Harvard Business Review study of 2,245 online enquiries; the “answered first, sold first” share is the Lead Connect survey; the data-entry discipline number is Salesforce’s annual State of Sales report; the new-versus-repeat cost ratio is the McKinsey estimate that has been quoted for a decade. We are not dropping links, but we are not inventing either — these are the same measured facts any serious funnel conversation is built on. They are also conservative: none of them works against you if you simply answer faster.

What disappears with a spreadsheet row

History. A client who bought twice is two unrelated rows in a sheet. There is no asset of “what we offered, why they refused, what they asked for next time”, because there is nowhere to keep it.

Ownership. A row with no responsible person is a row nobody finishes. In a CRM every lead has an owner and a deadline for the next step; in a sheet “someone will look” means no one will.

Automation. While the data lives in a file, no scenario can stand on it: not the payment reminder, not the invoice built from line items, not the 9:00 Telegram report for the owner. Exactly what a person does by hand in a sheet, a trigger does in a system — and the trigger does not get tired. Where the line runs between “assembled in no-code over a weekend” and “written as code” is covered separately in AI business automation.

Do the math honestly or it will not work. Put into the formula not “the average deal across the year” but what could actually have happened: enquiries you never answered within a day, “I’ll check later” notes that never came back, invoices that went out late. And do not multiply by a 100% conversion rate — use the one you have now.

Boxed or custom?

	Boxed (Bitrix24, HubSpot…)	Custom CRM
Start	from ₴40k + subscription	from $3,614 (₴150,000)
Process	you adapt to the box’s scenario	built around your process
Integrations (ERP, telephony, warehouse)	“sometimes via API, for extra money”	native, designed in from day one
Data	rented from the vendor	yours, in your database
After 3 years	subscription × 3 + limits	an asset you stopped paying rent on

The gap at the start looks bad for custom, so run three years on your own numbers. Boxed: ₴40,000 implementation plus licences — at ten users and ₴300 per seat that is ₴3,000 a month, ₴108,000 over three years, ₴148,000 all in, and that sum does not shrink even if you never change a thing. Custom: $3,614 (₴150,000) once, plus hosting and support. After that it depends on how much your process resembles “everyone else’s”: the more deviations, the more each one costs you in bending a box.

Three scenarios instead of “do we need a CRM”

Stay in the sheet. 10–15 deals a month, one or two sellers, stock that does not run out under your feet. Honestly: you do not need a system yet, you need discipline — one file, mandatory “next step” and “date” columns, an invoice template. That is a week of work, not a project.

Boxed, configured. Standard process (enquiry → quote → invoice → payment), typical integrations, a team up to 15. Starts around ₴40,000 with a subscription. The ceiling arrives in about half a year, when you need a field the vendor never planned for.

Custom. The process has specifics: stock, production, bonuses, many roles, 1C as the source of truth. Starts at $3,614 (₴150,000), in 2–3 week iterations. After three years it is your property, sellable together with the business.

Honestly: when a CRM will get in your way

If the team does not want to work in it, no CRM fixes that. A half-filled system is worse than a spreadsheet: it produces the appearance of order and numbers nobody believes. So we start not with development but with two questions: who exactly lost a deal last month because the data was not in a system, and what did it cost. If there is no answer, it is early yet.

Where to start: one free step that saves a month

What the custom CRM includes, and what it does not

Inside from $3,614 (₴150,000): pipelines mapped to your stages, roles and permissions, a client card with history, integrations with the channels enquiries actually arrive through, documents (invoice, act), an owner’s dashboard, hosting and handover of access. Contours you join later are costed separately: 1C/BAS from $964 (₴40,000) per contour, a non-standard connector from $241 (₴10,000), support after launch from $72 (₴3,000)/mo.

Not included, and impossible without you: filling the reference data, deciding “how it should work” in the owner’s place, training the team (we demonstrate, you drill). A CRM that was built and never handed to people is simply an expensive internal website.

Questions clients ask

Ready to price your case? Discuss a CRM → — free process audit, fixed estimate, from $3,614 (₴150,000). Related: choosing a CRM, lead response speed and integrating a site with 1C.