What an enquiry record shows, and what it cannot
Many guides on this site that mention a cost per customer, a lead-to-customer rateShare of leads created in a period that became paying customers. or the quality of enquiries assume the business knows who got in touch, by which route, and what happened next. The guide to unit economics for service businesses asks for enquiries, quotes, bookings and completed jobs counted by source from the business's own records. The diagnosis of enquiries from the wrong customers makes a spreadsheet, at first if need be, the one place where each enquiry's source and outcome are written. The diagnosis of marketing not generating leads wants the marketing report and what the people who answer enquiries have handled to agree. None shows how to keep that record, and the CRM guides assume a CRM already exists.
The gap is real for a business whose customers arrive by telephone, message, email, a visit or a recommendation. Analytics can record that a phone link was tapped, but the call itself happens off the site, as the guide to measuring calls and chats explains. Each record answers a different question:
| Record | What it can answer | What it cannot answer |
|---|---|---|
| Website analytics | Visits, form submissions, the tracked source of a visit | The call or conversation itself; what became of an enquiry; who it was |
| Phone log or call tracking | Calls by number or source, answered or missed | Whether the call became a booked job |
| Booking, order or invoice system | What was sold, when and for how much | Enquiries that were not sold, where the system does not record them |
| CRM | The whole path of each lead, where every route feeds it | Anything, where there is none or where only the web form lands in it |
| Enquiry record | Every enquiry with its route, its source and its outcome | Why the business was finally chosen, and what marketing caused |
The last cell matters. What customers say about how they found the business is a view of where attention started and cannot show what any channel caused. The self-reported attribution guide says so plainly and sends causal questions to the designs in the incrementality testing guide. The record's job is narrower: to make the business's own counts exist, so that later decisions can start from them.
Before you start
- Someone who answers enquiries and agrees to record each one. A record kept by the owner in a good week and by nobody in a bad week describes the good weeks. Name one person who owns the sheet and a person for each route who adds the row.
- A place the business controls. A sheet in a business-owned shared drive, listed in the access and ownership register that the new-business setup tutorial builds, not in one person's account or phone. It is First-Party DataData a business collects directly from its own customers, visitors and users., collected by the business itself.
- Access to every place an enquiry can land: the inbox behind the web form, the phone log or voicemail, the shared email address, each social and chat inbox, and a notebook or phone note where enquiries arrive in person.
- The business's other counts, such as the diary, the order or invoice list and the analytics report, which step eight checks the record against.
Set aside a sitting to build the sheet and agree its lists, a short moment for each enquiry as it arrives and a fixed slot each month to check and read it.
The steps
Write down what counts as an enquiry and every route it can arrive by
An enquiry is a real person asking for something the business sells: a price, a booking, a visit, a quote, an order. The signals around it are not enquiries. The calls guide separates taps on a phone link and chats opened, which it treats as diagnostics, from a qualified call or a chat that captured a request, which it treats as the enquiry, and it asks for the rule to be written down with an owner. Apply the same discipline here.
List every route in use: the web form, the telephone, the shared email address, each social inbox and chat tool, enquiries made in person at a counter, stall or event, enquiries passed on by a partner or another customer, and any platform or directory that forwards contacts. Then settle the awkward cases on paper. Does an existing customer asking for more work count? Does a supplier's call, a sales approach or a wrong number? The lead-to-customer rate entry treats which leads count as a definition to write down, and everyone who records enquiries and the person who reads the record must apply the same rule.
Choose one place, one owner and a rule for who may see it
Put the record in one sheet with one row for each enquiry and the same layout for every route. A second sheet for the telephone and a third for messages split the count across three places.
The sheet holds names and contact details, which are Personal DataAny information relating to an identified or identifiable living person, such as a name, email address, location data or an online identifier., so keep to what the business needs to reply and follow up, and limit access to the people who answer enquiries. The self-reported attribution guide applies the GDPR's data minimisation principle to free-text answers, and the guide to personal data in analytics names form fields as one way it leaks, so leave personal details out of analytics tools and build the monthly report from counts and codes rather than names. This is general practice, not legal advice.
Keep the record apart from any mailing list. The setup tutorial and the list-building and consent guide keep the two records separate, with consent held in its own consent recordThe stored evidence of a marketing permission: the channel, the wording the person saw, the date and the source of each opt-in, kept with the contact record. at its own sign-up.
Choose the columns, and add none that a decision will not use
Each column should answer a question that someone will ask at the monthly read. A column nobody reads is a column nobody fills in. A workable first layout:
Column Filled in by The question it answers Date received Whoever takes the enquiry Which month's cohort it belongs to Route Whoever takes the enquiry Which routes carry the enquiries, and whether each is covered Reference or name Whoever takes the enquiry Who to reply to; a code is enough in any report What they asked for Whoever takes the enquiry, from a short list Which jobs or products people want How they found the business, as told Whoever takes the enquiry Where attention came from, in the customer's words How they found the business, coded The person who owns the sheet Where attention came from, counted by category, with unclear as one of them Tracked source The form or the call system, where one exists Where the visit or the call came from, as tracking shows it First reply, date and time Whoever replies How long people wait Status The person handling it Where the enquiry is now Next step and date The person handling it Which enquiries are waiting on the business Outcome, reason and date The person handling it What became of it, why, and how old the enquiry was when it ended Value The person handling it, when won What each source is worth, where the figure is known The measurement plan guide has the plan name a Source of TruthThe system whose figure is treated as authoritative for a given number, such as billing for revenue or the CRM for pipeline, with other tools reconciled to it. for each number, and the measurement plan framework adds an owner and a cadence; for the count of enquiries this sheet is that source, and the booking, order or invoice system stays the source for what was sold.
Ask every enquirer how they found the business, and write down the answer as given
The self-reported attribution guide is the method; this step applies it to the sheet. Ask an open question, with no list to prompt the answer, in the same words each time and at the same moment on each route: at the end of a call, in the first reply to a message or on the form. The guide sets out which question to ask; pick one and keep it.
Record the answer in the person's own words, then code it later against a short codebook of channel-level categories with an explicit unclear category, so that the raw answer is kept and the coding can be redone. The guide lists the biases to read around. A change to the wording or the codebook is a break in the series, so date it in the measurement change log.
Capture the tracked source wherever the route can carry one
Some routes can carry a source and some cannot. A web form can: the campaign values in a tagged link, set by the conventions in the UTM and naming conventions guide and requested through the UTM link request SOP, can be copied into hidden fieldsA form field the visitor never sees that records values such as source, campaign and landing page, so each lead arrives in the CRM with its origin. on the form so that they arrive with the submission. The inbound lead handling playbook asks for the same: a source, a campaign and a landing page recorded with the lead. Have the form's notification carry those fields, so that the source is copied into the sheet with the enquiry.
A telephone number that differs by source is the call equivalent, and the calls guide describes Call TrackingAttributing phone calls to the marketing source that produced them, usually by showing different tracking numbers per source or visitor. and how to attribute each call. An enquiry in person, or by social message with no tagged link or prefilled first message, has no tracked source, and the answer to the source question is the only source there is. Keep the two columns side by side and do not let one overwrite the other: where they disagree, the disagreement is information.
Define a short list of statuses and the event that moves a row on
A status is a test, not a label. The lead stages and pipeline guide makes the point for a CRM: write what must be true of a record to be in a stage and what moves it on, and a count by stage means something. Keep the status list short. A first list might run from new, to answered, to quoted or visit booked where the job needs one, to booked or won, with a closed set at the end: not suitable, left at the price, chose another supplier, never reachable or no reply after the agreed follow-ups, withdrawn and not an enquiry.
Two points deserve care. Not ready and not a fit are different findings, as the pipeline guide explains for a CRM, and a sheet can keep the difference with a not-now status that carries a date to come back to. And a silent enquiry needs a rule, written down in advance, for when it counts as lost: the follow-ups the business will make and the time it will wait. The wrong-customers diagnosis offers a short set of outcome tags; as this tutorial's own method, tag a sample of real enquiries with the list and add or merge tags until it covers them.
Close every row with an outcome and a reason on the day it is known
Make the outcome and reason compulsory before a row may be closed, as the wrong-customers diagnosis does for the outcome, so that the record does not fill with enquiries that have no ending. Pick the reason from the list rather than writing a sentence, so that reasons can be counted, and allow a free note only for a reason the list does not yet hold; when a note recurs, promote it to the list.
Add the value when a job is won, as the quoted or invoiced figure, and leave it empty otherwise. Record the date the outcome became known as well as the date of the enquiry, because the gap between them is how long an enquiry takes to reach an outcome, and step nine reads recent months differently for that reason. A row left open without a next step and a date has no owner; as general practice, a regular sweep of open rows, on a rhythm the business writes down, keeps them honest.
Check the record against the business's own counts each month
Once a month, compare the record with the counts kept elsewhere, route by route. Enquiries recorded from the web form should agree, within a margin that can be explained, with the submissions the site's analytics reports, which the forms tracking guide counts with a lead event; calls entered in the sheet should reconcile with the phone log once calls that are not enquiries, by the written rule, are set aside; and every booked or won row should match a booking, a diary entry or an invoice. Check the other way too: each booking or invoice should have a row. A customer with no row is either a repeat or referred customer who did not go through enquiry, which the sheet should say, or an enquiry the record missed.
The test is the one the leads diagnosis sets: the report and the count kept by the people answering enquiries agree for a full month within a small margin that can be explained. Causes to look for include a route nobody records and test submissions or spam counted as enquiries; as general practice, also check for rows added late and for a form whose notification has stopped arriving. Fix the cause, note the fix with its date and reconcile again next month.
Read the record each month by source, by cohort and by reason
Group enquiries by the month they arrived, not the month they closed, and show each cohort's age, because the newest cohorts are provisional; the lead-to-customer rate entry sets out the method. Then read in this order:
Question What to read What it can feed Where do booked jobs come from? Booked rows by source as told and as tracked, each with its count Which sources to look at first when budget is next allocated, in the budget allocation guide Where do unsuitable enquiries come from? Not-suitable rows by source and reason The causes the wrong-customers diagnosis works through Where do enquiries leave at the price? Left-at-the-price rows by source The offer, and how and where the price is shown How long do people wait? Gap between receiving and first reply, by route The service levels in the inbound lead handling playbook What are people asking for? The asked-for list, counted Which lines to put marketing behind, in the guide to choosing products to promote What is stuck? Open rows by age and next step Follow-up, and capacity Put the count beside every rate, as the guide to reading small numbers asks, because a handful of enquiries swings a share. Read over longer periods where counts are low, compare with the business's own earlier months as a BaselineAn estimate of what a metric would have done without a given change, campaign or promotion, against which the effect is measured. rather than with anyone else's figures, and be wary of a source that looks strongest on the fewest rows. Where spend is known for a source, the unit economics guide sets a break-even and a target cost per booked job to read it against, and the CPA calculator does the cost arithmetic.
Use what the record shows, and decide when it has outgrown the sheet
Change one thing at a time, log it and read the record after the change against the record before it. The record is evidence for decisions, not a replacement for judgement: it can show that unsuitable enquiries cluster in one source and leave what to do about it to the targeting, the listing or the terms.
Where enquiries come from paid clicks, outcomes can go back to the platform. That needs the click identifier captured on the form, which the guide to linking the CRM to ads and analytics describes, a consent check and an Offline Conversion ImportUploading outcomes recorded outside the website, such as closed deals, to an ad platform. run as a routine, which the offline conversion upload SOP sets out. The connecting guide lists a spreadsheet upload among the routes; as general practice, a sheet kept with the same fields can stand in for the export the SOP starts from until there is a CRM.
As general practice, treat these as signs that the record has outgrown a sheet: rows are missed or duplicated because several people edit at once, enquiries need routing to different people, follow-ups depend on someone remembering to look, outcomes should reach the ad platforms without retyping, or the monthly check takes longer than the decisions it informs. The CRM fundamentals guide describes what a CRMThe system that stores an organisation's contacts, companies, deals and interactions, and the processes built around it. is for, and the moving to a new CRM tutorial covers carrying history from one CRM to another. The columns, the status list and the written rule for what counts as an enquiry move with the record, so they are worth getting right now.
Where the record leads
The record is an input rather than a finished analysis. Once a business has a record and a CRM beside it, the scored lead management audit checks the loop between marketing and what became of the enquiries, and the core metrics guide shows how a count of new customers feeds customer acquisition cost. The fundamentals hub lists the other practical tutorials that put the basics to work.
A mosaic studio reads its first quarter of enquiries (illustrative)
Dratolmi Mosaics reads its first quarter of enquiries
Dratolmi Mosaics, an invented studio that makes mosaic panels and splashbacks to commission, takes enquiries by web form, telephone, email, social message and at a craft fair stall. Its owner and a part-time assistant answer them. Its analytics report showed 21 enquiries in the quarter, all from the web form; once the owner builds a sheet on the lines above and the assistant adds a row for every enquiry, the record shows 64: 21 by form, 17 by telephone, 11 by email, 9 by social message and 6 at the stall.
The source each person gave is read beside the outcomes. Four enquiries have no usable answer and are coded unclear.
| Source, as told | Enquiries | Booked | Not suitable | Left at the price | Never reachable | Still open |
|---|---|---|---|---|---|---|
| A search result | 19 | 5 | 8 | 2 | 1 | 3 |
| A friend or past customer | 16 | 8 | 1 | 3 | 0 | 4 |
| A social post | 12 | 3 | 1 | 5 | 1 | 2 |
| The craft fair | 8 | 2 | 0 | 3 | 1 | 2 |
| A directory page | 5 | 0 | 4 | 0 | 0 | 1 |
| Unclear | 4 | 1 | 0 | 0 | 2 | 1 |
| Total | 64 | 19 | 14 | 13 | 5 | 13 |
Of the 21 form enquiries, tracking credited 13 to direct visits, and 6 of those 13 named a friend or past customer, so the studio reads the two columns together rather than trusting either alone.
Thirteen rows are still open, so the studio treats the booked counts as a floor, reads the first month's cohort, which is older, before the third and moves no money on 64 rows. Of the 14 rows marked not suitable, 8 came from search and 4 from the directory page, a source of only 5 rows that the studio reads as a question and not a finding. The reasons are an install address outside the area the studio serves and jobs below its minimum, so before touching any targeting it states its service area and its smallest job on the directory listing and the landing page, as one change to its visible terms.
The social post group has the largest count left at the price, 5 of its 12 rows, so the owner looks at what those posts show without saying what a commission costs. The friend or past customer group has the most booked rows, 8 of its 16, and nothing in the record says whether those people would have come anyway, so the owner notes it for later and claims no result. The row-by-row check also finds 3 bookings with no enquiry row, all repeat customers, which the owner records as such.
Next quarter's record will show how the share of not suitable rows moved after that change. The source question keeps its wording, and the studio dates the change to its visible terms in the measurement change log.
Common mistakes
- Recording only the web form. The routes that analytics cannot see are the ones the record exists for. A sheet that copies the form notifications has rebuilt the analytics report.
- Letting the tracked source overwrite what the person said, or the reverse. Keep both columns; their disagreement is a finding.
- Closing rows without an outcome or a reason. The sheet then shows a pile of enquiries and no endings, and the reasons that would change targeting or the offer are not counted.
- Setting this month's customers against this month's enquiries. Group by the month enquiries arrived and show each cohort's age; the newest are provisional.
- Treating the record as a mailing list. An enquiry is not consent to marketing, and the two records live apart.
- Reading it as proof of what a channel caused. The record says where attention came from and what became of it. Whether a channel added enquiries that would not otherwise have come is a question for a test.
