Spreadsheet or CRM? When an Online Shop Should Change Its Workflow

A Tale of Two Shops

Imagine two fictional Thai online shops, both selling handmade ceramics. The first, run by a single founder named Maliwan, has around 40 returning customers and handles every order herself. She keeps a Google Sheet with four columns: customer name, phone, last order date, and glaze-preference notes. It works. She opens the file in the morning, scans down the 40 rows, and knows exactly who needs a follow-up. A solo owner can maintain a consistent process, but still needs to check for missing records and incorrect entries.

The second shop, Dao Ceramics, has grown. Dao hired two part-time sales reps — Krit and Pim — to handle Line and Facebook inquiries separately. They share one Google Sheet, but Krit updated a customer's status to 'Quoted' on Monday while Pim, working from a downloaded copy she had saved on her laptop, overwrote that entry with 'New Lead' on Wednesday. The customer received two conflicting messages. A week later, Pim left, and no one could find the conversation history she had kept in a separate chat log on her phone.

These shops frame a workflow question: can the current process support the people and interactions it carries? This guide examines when a spreadsheet can remain suitable and which observed problems justify comparing improved spreadsheet controls with a CRM trial. Consider configuration, team habits and costs before choosing.

Why the Spreadsheet Deserves a Fair Hearing

There is a tendency in business content to treat spreadsheets as primitive tools that serious operators should discard quickly. That framing is unhelpful. Spreadsheets are free or near-free, require almost no onboarding, and are flexible enough to model almost any structure a small team can imagine. For a shop with one person managing relationships, these qualities are genuinely valuable rather than merely tolerable.

Vendor comparison content from HubSpot positions spreadsheets as appropriate for teams of fewer than five people tracking under 50 active opportunities. Salesforce's equivalent materials describe a rough tipping point at around 50 active accounts or more than two people regularly editing the same file. It is worth being honest that these figures come from vendor marketing content rather than controlled research, so they should be treated as plausible orientation points rather than precise thresholds. What they capture correctly is the underlying logic: the failure modes of a spreadsheet are almost all social and concurrent — they appear when multiple people try to use the same static file as a live, shared system.

A spreadsheet also remains the right tool for specific tasks even inside a shop that has already adopted a CRM. One-off data cleaning, scenario modeling, financial projections, and bulk CSV preparation before an import are all tasks where a spreadsheet's formula flexibility and direct cell manipulation are genuinely superior to most CRM interfaces. The decision is not binary in practice; it is about which tool is the primary source of truth for customer relationships.

Sources: HubSpot: Advantages and Disadvantages of Using Spreadsheets for Business · Salesforce: Spreadsheet vs CRM — 5 Signs Your Small Business Has Outgrown Spreadsheets

When a Spreadsheet Still Suffices

Before cataloguing the triggers that push a shop toward a CRM, it is worth being specific about the conditions under which staying with a spreadsheet is a reasonable, defensible choice — not just a cost-cutting compromise.

A solo operator with a stable customer list may have fewer handoffs to manage, but still needs to check errors, backups and record completeness. A spreadsheet may also suit bounded reference or planning work, such as tracking competitor prices or supplier contacts, when its fields and update responsibilities are clear. Choose from the actual task rather than assuming that one user guarantees correctness.

Financial modeling is another area where spreadsheets hold a structural advantage. Calculating margin scenarios, building a cash-flow projection, or modeling commission structures requires formula flexibility and direct numerical manipulation that most CRM platforms do not offer. Shops that need both a CRM for customer relationships and a spreadsheet for financial planning are not in contradiction — they are using each tool for what it does best.

HubSpot's example about fewer than 50 active opportunities refers to sales opportunities, not a universal customer or lead cutoff. Treat vendor thresholds as contextual examples. If one person can maintain the required records and follow-ups, a spreadsheet may remain suitable; test the workflow before buying another tool. The four cases below are editorial comparison prompts, not numerical suitability estimates or measured outcomes.

Trigger 1: Duplicate Edits and Version Conflicts

Conflicting copies are a concrete workflow problem: one rep uses a downloaded file while another updates the shared source, and the team does not reconcile which change is current. This is a reason to investigate that process, not proof that every spreadsheet workflow needs replacement.

In the fictional Dao Ceramics case, Krit updates the shared source while Pim uses an old local copy. Google Sheets supports restricting edits to selected sheets or ranges. Check whether those controls, one agreed working file and explicit ownership solve the observed problem. Range protection is not a security measure; it does not prevent every copy or export. Review sharing and access separately.

Vendor articles cite differing spreadsheet-error percentages — HubSpot's source references a study finding 94% of business decision-making spreadsheets contain errors, while Salesforce's source puts the figure at close to 90% — without sufficient methodological detail to apply those rates directly to any particular shop. Treat them as vendor context illustrating why error rates tend to grow with data volume, not as your team's measured error rate or a guaranteed increase as editors are added. Record actual overwritten statuses, duplicate contacts and recovery time before comparing alternatives.

A CRM can provide shared records, ownership, permissions and change history when the selected product, plan and configuration support them. These features still need a working process. Exports, offline copies and stale imports may remain possible; test concurrent changes instead of assuming the purchase eliminates conflicting data.

Sources: HubSpot: Advantages and Disadvantages of Using Spreadsheets for Business · Salesforce: Spreadsheet vs CRM — 5 Signs Your Small Business Has Outgrown Spreadsheets · HubSpot: How to Build a CRM in Google Sheets · Google Docs Editors Help: Protect sheets and ranges

Trigger 2: Missed Callbacks and Unclear Ownership

The second trigger is subtler but equally damaging: follow-ups that were supposed to happen but did not, because no mechanism existed to make the responsibility unambiguous. In a spreadsheet, ownership is typically expressed as a name in a column, a colour fill, or an informal agreement about which rows belong to which rep. None of these mechanisms generate a reminder, escalate an overdue task, or make it visible to a manager that a high-value lead has not been contacted in ten days.

Consider a hypothetical scenario: Krit at Dao Ceramics adds a note in column G — 'Call back Friday' — next to a customer who expressed interest in a bulk order. He falls ill on Thursday. Pim scans the sheet on Friday but does not see the note because she is filtering by her own name in column C. The bulk-order customer is never called back. This is not a failure of effort or intention; it is a structural failure of the tool.

Salesforce identifies broken follow-ups as one of five operational triggers for switching to a CRM, alongside version conflicts, slow lead response, vanishing customer histories, and forecasting blind spots. It is worth noting that these are qualitative signals from vendor marketing content, not quantified thresholds with associated revenue data. But the pattern they describe — tasks that depend on someone remembering to look at the right row at the right time — is structurally fragile in a way that a CRM's task assignment, due-date notifications, and pipeline stage automation directly address.

The practical test for this trigger: track whether customers or prospects have gone without follow-up because it was unclear whose responsibility it was, or because a reminder lived only in one person's memory or private notes. If this pattern recurs across multiple instances within a short period, the ownership problem may have become a recurring workflow defect rather than an isolated oversight. Log these incidents during a trial observation period and use the count to inform the decision.

Sources: Salesforce: Spreadsheet vs CRM — 5 Signs Your Small Business Has Outgrown Spreadsheets

Trigger 3: Customer History Walks Out the Door

The third trigger is the one that tends to produce the most acute regret: a rep leaves, and their entire knowledge of a customer's preferences, conversation history, and negotiation context disappears with them. In a spreadsheet environment, interaction notes often live in a personal chat log, a private inbox, a locally saved file, or a column that the departing rep alone maintained. When they leave, the shop faces a high-value customer who remembers every detail of three previous conversations — and a new contact who knows nothing.

Salesforce's vendor content describes this scenario as 'vanishing customer histories' — a salesperson's departure taking account history and client notes to a personal drive. The same pattern appears in multiple Salesforce materials, including both their small-business blog and their SMB implementation readiness guide, which suggests it reflects a genuinely recurring operational concern rather than an edge case.

A CRM can centralize interactions that the team records or that enabled features and integrations capture. Whether calls, messages, notes and stage changes are retained and visible depends on product capabilities, configuration, permissions and logging habits. Before a staff handoff, verify that the required history is stored in the shared record and that the next authorised colleague can read it. Buying a CRM alone does not guarantee complete context.

The practical test: if your most active sales rep left tomorrow, how much of your customer relationship history would survive? Assess this honestly by checking where interaction notes currently live — shared system, personal device, or private chat. If the majority exists only outside a shared system, this trigger warrants investigation.

Sources: Salesforce: Spreadsheet vs CRM — 5 Signs Your Small Business Has Outgrown Spreadsheets · Salesforce: CRM Implementation Readiness for SMBs

Threshold Indicators: How to Read the Signals Together

No single threshold number determines when to switch. The figures cited by vendor content carry different objects: HubSpot's spreadsheet-vs-CRM comparison table sets a rough limit of under 50 active opportunities and fewer than five people; HubSpot's Google Sheets CRM guide suggests upgrading when managing more than 100 active leads or more than 500 total customer records; Salesforce's small-business blog cites a tipping point at around 50 active accounts or more than two team members editing the same file. Each figure refers to a different measure and comes from vendor marketing content rather than peer-reviewed research. Use them as orientation prompts rather than precise cutoffs.

Salesforce's readiness content notes that managing 500 customers across multiple departments makes manual spreadsheets harder, and that the average SMB uses seven different applications, with 46% of SMB respondents feeling overwhelmed by that fragmentation — figures from Salesforce's SMB Trends Report. These observations describe a particular scale and context; they should not be generalised into a universal rule. A shop managing 30 custom wholesale accounts may need a CRM sooner than a shop with 150 transactional retail customers who rarely require follow-up.

A more reliable approach is to evaluate the three triggers as a composite signal. Use the table below as a self-assessment. Each row represents an observable condition; treat the right-hand columns as diagnostic prompts rather than definitive rulings — the presence of several 'CRM likely needed' conditions is a stronger indicator than any single headcount figure.

Sources: HubSpot: Advantages and Disadvantages of Using Spreadsheets for Business · HubSpot: How to Build a CRM in Google Sheets · Salesforce: Spreadsheet vs CRM — 5 Signs Your Small Business Has Outgrown Spreadsheets · Salesforce: CRM Implementation Readiness for SMBs

What the Transition Actually Looks Like

Understanding when to switch is only useful if the transition is practical. Salesforce's migration guidance suggests allowing one to three months for a small CRM migration, with the timeline depending on data complexity, workflow mapping, and team size. For a small Thai online shop moving from a Google Sheet to a CRM, the realistic path breaks into three phases.

The first phase is data preparation; estimate its duration from your actual files and team capacity. Export the current spreadsheet as a CSV file. Before importing anything, clean the data: remove duplicate rows (check by email address, phone number, or a combination of name and company), standardise field formats (phone numbers in one format, date columns in one convention, status labels consistent throughout), and decide which records to migrate versus archive. Salesforce's readiness guide is explicit that inconsistencies in the source data follow you into the CRM — cleaning before import is not optional.

For field mapping, choose the destination object and check its required fields, identifiers, types and import permissions. A contact and a deal are different records; supported combined or separate imports need explicit associations. Put purchase dates in an appropriate order field or custom Last Order Date, not an activity timestamp. Create needed custom fields, test a representative small sample and read the saved records back before the full import.

The third phase is team adoption, which is often underestimated. A CRM is only as good as the interaction records added to it. If reps continue logging notes in personal chats and only updating the CRM occasionally, the system becomes as unreliable as the spreadsheet it replaced. Clear standards — what gets logged, when, by whom — established in the first week of use determine whether the tool delivers on its structural promise.

For shops using CasperBrain for retention analysis, the same CSV export discipline applies: clean, deduplicated customer records with consistent order identifiers are the foundation for any downstream analysis. CasperBrain supports authorised CSV exports, and field visibility follows each user's permissions — so investing in clean, well-structured data at the transition point pays forward into every subsequent analytical step.

Sources: Salesforce: CRM Implementation Readiness for SMBs · HubSpot: How to Build a CRM in Google Sheets · HubSpot Knowledge Base: Import records for multiple objects

A Worked Example: Mapping a Sheet to CRM Fields

This fictional Baan Craft example maps eight columns to one contact object. It is a planning example, not a universal import template. Create the named custom fields with the stated types, or select equivalent supported fields. Order/deal history needs that CRM's separate objects and associations; a contact summary does not create it. The name, example.com email and XXX phone are fictional teaching placeholders. Do not upload this sample as real customer data.

Sources: HubSpot Knowledge Base: Import records for multiple objects

Handling Buddhist Era Dates in Your CSV

First confirm modern Buddhist Era text dates in day/month/year order: 15/3/2567 corresponds to 2024-03-15. A CRM may reject or misinterpret unconverted text; do not assume its parser recognises the era. Do not subtract 543 from an already-correct CE date or spreadsheet date serial.

Keep A2 as original text, such as 15/3/2567, with empty helper cells B2:E2. In Excel for Microsoft 365 or Excel 2024, enter =TEXTSPLIT(A2,"/") in B2 to spill day, month and year into B2, C2 and D2. In E2 enter =DATE(VALUE(D2)-543,VALUE(C2),VALUE(B2)) and format it as yyyy-mm-dd: the result is 2024-03-15. Splitting on / also handles 5/3/2567 and 05/03/2567 without fixed-width assumptions. Without TEXTSPLIT, use Text to Columns with / and preserve the parts as text, then apply the same DATE formula. Validate day/month order and known examples first: DATE can roll an invalid date into another month. Formula argument separators depend on spreadsheet locale.

Keep a source backup. Copy checked E-column results and paste values into the destination date column before removing helpers, so formulas do not lose references. Export a separate CSV and inspect its raw yyyy-mm-dd values, not only the spreadsheet display. Verify dates after a trial import and retain the original until reconciliation.

Sources: Microsoft Support: TEXTSPLIT function · Microsoft Support: DATE function

Limitations of a CRM — Honest Caveats

A CRM is not a solution to every operational problem, and it is worth being direct about the limitations before a shop owner commits to the transition.

First, a CRM does not automatically populate itself. The version-conflict and ownership problems disappear only if reps actually log interactions in the system. A CRM with incomplete records is not better than a well-maintained spreadsheet — it is worse, because it creates a false sense of completeness. Adoption discipline matters more than the software choice.

Second, the threshold figures discussed throughout this guide — around 50 active opportunities or accounts, around 100 active leads, two-plus reps — are vendor-sourced guidelines with different objects and contexts, not empirical benchmarks. A shop with 60 accounts but a very stable team and clear informal ownership conventions may find that a spreadsheet still works adequately. A shop with 30 accounts but high staff turnover and chaotic follow-up practices may need a CRM urgently. Read the triggers as diagnostic questions, not as bright-line rules.

Third, CRM platforms vary considerably in their setup complexity, cost structure, and feature sets. When evaluating options, check separately whether a platform offers a Thai-language interface and whether it supports the communication channels your team actually uses — for example, by reviewing each vendor's integration directory or documentation. These are platform-specific factors that require hands-on evaluation alongside the workflow maturity assessment this guide addresses.

Fourth, importing data from a spreadsheet into a CRM does not automatically improve data quality. Duplicate records, inconsistent status labels, and missing email addresses in the source sheet become duplicate records, inconsistent labels, and missing fields in the CRM. Cleaning before importing is essential, not optional.

Finally, for shops using retention analysis tools such as CasperBrain alongside a CRM, it is important to understand that these are separate products with separate data flows. CasperBrain supports authorised CSV exports for analysis, and field visibility follows each user's permissions within CasperBrain. Planning the data pipeline — what gets exported, when, and in what format — is a deliberate step that needs to be designed, not assumed.

Sources: Salesforce: CRM Implementation Readiness for SMBs · Salesforce: CRM Template for Small Business

Making the Decision: A Practical Checklist

The decision to stay with a spreadsheet or move to a CRM reduces to a small number of honest questions. Work through them in order.

Is more than one person regularly editing the same customer file? If yes, the version-conflict risk is active. Move to the next question to assess urgency; do not wait for a damaging incident to confirm the problem.

Have follow-ups been missed — or nearly missed — because it was not clear who owned them? If this pattern recurs within a short observation period, the ownership problem may be structural. Track the incidents explicitly rather than relying on memory: a note in a column is not a task assignment.

Would a key rep's departure take significant customer context with it? Assess this by checking where interaction notes currently live. If the majority exists only on personal devices or in private chats rather than a shared system, the history-loss risk is real and ongoing.

Are there more than roughly 50 active opportunities or accounts being managed across the team — or more than 100 active leads? These are orientation figures from different vendor sources, not a single hard rule. Use them as prompts to re-examine the first three questions at greater urgency, not as definitive cutoffs.

If all four answers are no, a well-maintained spreadsheet may remain suitable. Several yes answers are a prompt to investigate, not a validated buying threshold. Compare improved spreadsheet controls with a small CRM trial using the same team, tasks and observation period. Record missed follow-ups, conflicting edits, recovery time and costs, then choose according to actual results and what your team can maintain.

Sources and editorial method

Examples are fictional and contain no customer records. Prepared by the CasperBrain team with AI drafting assistance; facts and language are checked before publication.

• HubSpot: Advantages and Disadvantages of Using Spreadsheets for Business
• HubSpot: How to Build a CRM in Google Sheets
• Salesforce: Spreadsheet vs CRM — 5 Signs Your Small Business Has Outgrown Spreadsheets
• Salesforce: CRM Implementation Readiness for SMBs
• Salesforce: CRM Template for Small Business
• Microsoft Support: TEXTSPLIT function
• Microsoft Support: DATE function
• HubSpot Knowledge Base: Import records for multiple objects
• Google Docs Editors Help: Protect sheets and ranges