You can finally say which marketing produced the revenue.
Gully Sales connects your campaigns, enquiries and closed deals into one traceable record, so every rupee of marketing spend is judged on the business it brought in, not on the activity it created.
- Every enquiry arrives carrying the source, campaign and touch that produced it.
- One agreed credit rule, so marketing and sales stop arguing over the numbers.
- Spend decisions made on traced revenue, not on clicks, likes and impressions.
Gully Sales Private Limited builds attribution for Indian SMBs, and states plainly what the data can and cannot prove.
In one paragraph
What is Marketing Attribution Services?
Marketing attribution is the method that connects each enquiry, opportunity and closed sale back to the campaigns, channels and touches that produced it. Gully Sales maps your real buying journey, fixes the source data at the point of capture, links it to your CRM, and builds attribution views with honest limits and written rules for acting on them.
The problem
Everyone has an opinion on what is working, and nobody has the record.
You spend on several things at once. Search, social, a listing site, an exhibition, a referral push, and somebody quietly calling old leads. Enquiries arrive. Some become customers. When you ask which activity produced them, the answers come from memory and confidence rather than from a record. So the budget follows whatever was discussed most recently, and the activity that quietly brings your better customers stays underfunded.
You will recognise it as
- Half your enquiries sit in the CRM as website, direct or blank, which tells you nothing you can act on.
- Your ad platforms together claim more leads than your CRM actually received that month.
- Sales says the leads are poor, marketing says the leads are not followed up, and neither side can show it.
- A customer who first found you on search and then messaged on WhatsApp is credited entirely to WhatsApp.
- You know your overall cost per lead, but not what a lead from each source is worth once it closes.
- Every review of spend ends with the same sentence: we think this one is working.
What it costs the business
- Budget moves on the strength of the argument in the room, so channels that build demand slowly are cut before they mature.
- Agencies and ad platforms are judged mainly on the numbers they report about themselves.
- Sales stops trusting the marketing report and keeps a private version of the truth in a spreadsheet.
- The activity that brings your highest value customers stays invisible, so it never gets more money.
Why it persists. Attribution fails at the point of capture, not at the point of reporting. The source is lost in a phone call nobody logged, a WhatsApp enquiry with no origin, a form that does not pass the campaign through, a visitor who saw a hoarding. Fixing that needs small changes in several places at once: forms, phone handling, CRM fields and sales habit. In a small business no single person owns all four, so the report gets rebuilt again instead, and a report cannot recover data that was never captured.
If it stays unresolved. Spend keeps growing while confidence in it shrinks. Decisions get postponed because nobody wants to defend a number in front of the owner. Then a slow quarter arrives, and the first thing cut is the marketing nobody can defend, which is often the channel that was quietly filling the pipeline for the quarter after.
What changes
You get a traced record of demand, and rules for what to do with it.
In the first weeks
- You can see, for the current month, where every recorded enquiry actually came from.
- The gap between what the platforms claim and what your CRM holds is measured and explained.
In how the work runs
- Every enquiry route, whether form, call, WhatsApp, walk-in or referral, captures a source the same way.
- One source and campaign vocabulary used by your team, your agency and your CRM together.
- Sales records how the deal was found, in a field that takes five seconds to fill.
In sales and marketing
- Budget conversations move from opinion to a shared view of traced pipeline and revenue.
- Cost per qualified lead and per won customer can be compared by source, not only by platform.
- Weak spend can be reduced with evidence, and moved elsewhere under a written rule.
In what management can see
- Revenue is visible by first touch and by last touch, so discovery and closing both get their due.
- Long buying cycles show up as journeys rather than as isolated last clicks.
Over the longer term
- The credit rule survives staff changes, because it is written down and applied the same way each month.
- Each year of clean source data makes the following year's planning less of a guess.
Gully Sales controls the journey model, the tracking, the CRM connection, the attribution views and the decision rules. What follows from them, such as better allocation, lower acquisition cost or more revenue, depends on your market, your offer and how consistently your team applies the rules.
Who it is for
This is for you if you spend on more than one thing at a time.
The businesses it suits
- Businesses spending on two or more channels every month with no agreed way to compare them.
- Owners asked to approve a marketing budget without a record of what the last one produced.
- Marketing heads whose reporting is challenged by the sales team in every review.
- Companies with a long buying cycle where the first touch and the closing touch are months apart.
- Businesses that take enquiries by phone and WhatsApp as well as through a form.
- Teams already running a CRM that holds deal values but no reliable source field.
What usually prompts the call
- You are about to increase marketing spend and cannot say what the current spend returned.
- Two agencies or two ad platforms are claiming credit for the same leads.
- A channel is being considered for closure and nobody can prove the case either way.
- You have just moved to a new CRM and want the source data to be right from the start.
- An investor or a board member has asked for cost of acquisition by channel.
What Gully Sales does
The work, component by component.
Buying journey model
We map how your customers actually reach you: where they first hear of you, what they check next, who they ask, how many weeks pass before they enquire, and which step is never recorded anywhere. The map is built from your own enquiry records, your sales team's account of recent deals, and short conversations with recent customers where you allow them.
- Why it matters:
- A model built on how your buyers really behave is what makes the later numbers arguable in good faith rather than merely produced.
- You receive:
- A one-page buying journey map with typical touch counts, elapsed time, and the points where source data is lost today.
- Business value:
- You stop measuring a journey that does not exist, and you know which gap to close first.
Source and campaign taxonomy
We agree one vocabulary for sources, mediums and campaigns and write it as a list your team, your agency and your CRM all draw from. Every offline route gets a defined value too, including exhibitions, hoardings, dealer references, existing customers and outbound calling, so nothing has to be entered as other.
- Why it matters:
- Most attribution failures are naming failures. Three spellings of the same channel make a report unreadable long before any model is applied to it.
- You receive:
- A source, medium and campaign taxonomy sheet with allowed values, naming rules and worked examples.
- Business value:
- Two years from now, this month's campaigns will still be comparable with next year's.
Tracking implementation
We put the capture in place: link tagging for every paid, organic, email and offline destination, form fields that carry the source through to the CRM, calls and WhatsApp enquiries logged against a source, a short self-reported question on the enquiry form, and consent handled the way your privacy notice states.
- Why it matters:
- Data that is not captured at the moment of enquiry cannot be recovered later by any amount of analysis.
- You receive:
- Tagged link sheet, form specification, call and WhatsApp logging routine, and a completed tracking test record.
- Business value:
- The share of enquiries arriving with a usable source rises inside the first reporting cycle.
CRM and revenue connection
We connect capture to money: source fields on the lead and on the opportunity, inheritance rules so the source survives conversion, stage and value discipline, and the join that lets a closed deal be traced back to its first and last recorded touch. Where your CRM cannot hold a field, we design the workaround and document it.
- Why it matters:
- Attribution that stops at the lead measures interest. Attribution that reaches the invoice measures the business.
- You receive:
- CRM field specification, source inheritance rules from lead to opportunity, and a data quality checklist.
- Business value:
- Revenue, not enquiry count, becomes the number your channels are judged on each month.
Attribution views and credit rules
We build the views your decisions need: first touch for discovery, last touch for closing, a position-based or linear view for long journeys, and self-reported source as an independent check on all of them. Each view is defined in writing, with the question it answers and the question it must never be used for.
- Why it matters:
- One model cannot serve every decision, and settling on a single model quietly hides the channels that start journeys rather than finish them.
- You receive:
- An attribution report pack with first touch, last touch, multi-touch and self-reported views, each defined in writing.
- Business value:
- Discovery channels and closing channels are both visible, so neither is cut by accident.
Limits and caveats register
We write down what your attribution cannot see and in which direction it errs: walk-ins, word of mouth, conversations at a trade counter, blocked cookies, app to browser jumps, shared office connections and enquiries forwarded by a colleague. Each limit is stated with a way to sense-check it.
- Why it matters:
- A model presented without its limits is over-trusted for one quarter and then abandoned entirely when somebody finds a hole in it.
- You receive:
- A caveats register listing each known blind spot, its likely effect on the numbers, and how to check it.
- Business value:
- Your team argues about decisions instead of arguing about whether the data is real.
Decision rules and reallocation routine
We agree in advance what you will do when the numbers say something: the minimum volume before a source may be judged, the period it is judged over, the threshold at which spend is reduced or increased, and who signs that off. The rules are written into your monthly review agenda so they are used rather than remembered.
- Why it matters:
- Attribution only pays for itself at the moment a budget actually moves because of it.
- You receive:
- A written decision rule set, thresholds by source, and a monthly reallocation agenda with named owners.
- Business value:
- The monthly report ends in a decision rather than in a discussion that repeats itself.
What you will have at the end.
- Buying journey map with typical touch counts, elapsed time and the points where source data is lost today.
- Source, medium and campaign taxonomy sheet with allowed values, naming rules and worked examples.
- Tagged link sheet covering every paid, organic, email and offline destination you use.
- Form and enquiry capture specification, including the self-reported source question.
- Call and WhatsApp enquiry logging routine, written for the person who answers them.
- CRM field specification and source inheritance rules from lead to opportunity to won deal.
- Attribution report pack with first touch, last touch, multi-touch and self-reported views.
- An anonymised sample report, then the same pack populated with your first traced month.
- Caveats register naming each blind spot and its likely effect on the numbers.
- Platform to CRM reconciliation sheet explaining where the counts differ and why.
- Decision rules with volume thresholds, judging periods and named approvers.
- One training session and a recorded walkthrough for whoever runs the monthly review.
How it runs
The engagement, step by step.
- 1
Enquiry and data audit
We take a sample of recent enquiries and closed deals and trace each one backwards as far as your records allow. That tells us what share of your demand is traceable today, where the trail breaks, and how far the platform numbers sit from what your CRM actually holds.
- You provide:
- A CRM export or enquiry records for the last six to twelve months, and access to your ad and analytics accounts.
- We produce:
- A traceability baseline stating the share of enquiries and revenue that can be traced today, with the reason for each gap.
- Done when:
- You have agreed the baseline figure this engagement will be measured against.
- 2
Journey mapping and model choice
We map the real buying journey with your sales team, then choose which attribution views suit it. A two-week purchase and a nine-month capital purchase need different credit rules, and that is settled in writing before anything is built.
- You provide:
- Time with the sales lead, and permission to speak to two or three recent customers where that is appropriate.
- We produce:
- The journey map, the chosen attribution views, and the written question each view exists to answer.
- Done when:
- You have approved the credit rules in writing before any tracking work begins.
- 3
Taxonomy and tracking build
We write the source vocabulary, tag every destination, specify the forms, and set up call and WhatsApp capture. Running campaigns are renamed to the standard where that can be done without losing history, and the changeover date is recorded so old and new data are never silently mixed.
- You provide:
- Website and form access or your developer, ad account access, and the phone numbers used in marketing.
- We produce:
- Taxonomy sheet, tagged link sheet, form specification, capture routines and a completed tracking test record.
- Done when:
- A test enquiry sent through each route arrives carrying its correct source.
- 4
CRM connection
We add or correct the source fields, set the inheritance rules so a lead's source survives conversion into an opportunity, and check that deal values and close dates are entered consistently enough to be reported on at all.
- You provide:
- CRM administrator access and time with whoever maintains the pipeline day to day.
- We produce:
- Field specification, inheritance rules, a data quality checklist and the corrected CRM configuration.
- Done when:
- A closed deal from the sample can be traced back to its first and last recorded touch.
- 5
Reporting pack and caveats
We build the attribution report pack, populate it with your first traced month, and write the caveats register beside it. The pack is walked through with the people who will use it, in the meeting where it will actually be used.
- You provide:
- Attendance from marketing, from sales, and from whoever approves the budget.
- We produce:
- The report pack, the first populated month, the caveats register and the reconciliation sheet.
- Done when:
- Your team can read the pack and say what each view does and does not prove.
- 6
Decision rules and handover
We agree the thresholds, the judging periods and the approvers, write them into your monthly review agenda, and train the person who will produce the pack each month. Anything still unresolved is listed with an owner and a date rather than left implied.
- You provide:
- A decision on who owns the monthly pack and who approves a change in spend.
- We produce:
- The decision rule set, the review agenda, a training session and a recorded walkthrough.
- Done when:
- One full monthly review has been run by your own team, with us only observing.
- 7
Review after two cycles
Once two reporting cycles have passed we look at what the model actually changed. Sources with too little volume to judge are pooled, definitions that caused confusion are rewritten, and any blind spot that turned out to matter more than expected is addressed.
- You provide:
- Two months of reports and a short conversation about what was hard to use.
- We produce:
- A revision of the taxonomy, the views and the rules, plus a short note on what to fix next.
- Done when:
- The pack has survived two real budget conversations and been adjusted for them.
Ways to work with us
Choose the level of attribution your business can actually maintain.
Attribution audit
A short diagnostic that measures how much of your demand is traceable today, where the trail breaks and by how much the platform numbers overstate it. It ends with a prioritised fix list you can act on with your own team.
Attribution build
The full engagement: journey model, taxonomy, tracking, CRM connection, report pack, caveats register and decision rules, built alongside your team and handed over documented.
Build with reporting support
The build, after which we produce and present the monthly attribution pack for an agreed period while your team learns to run it, then hand the pack over completely.
Long-cycle attribution review
For businesses whose deals close over many months: a quarterly analysis that follows each cohort of enquiries through to revenue, instead of judging a channel by the month its money was spent in.
Why Gully Sales
What you are actually choosing when you choose us.
We build for the enquiry routes Indian businesses actually use.
Phone calls, WhatsApp messages, walk-ins and dealer references carry a large share of demand here. A setup that counts only web forms will misread your business, and usually in favour of whatever is easiest to track.
We state the limits in writing.
Every model we hand over arrives with a register of what it cannot see and which way that bends the numbers, so your team can rely on the pack without over-trusting it in the room.
We connect attribution to money, not to clicks.
The work is finished when a closed deal in your CRM can be traced back to a source, and cost per won customer can be compared across channels on definitions everyone has agreed.
We also run the marketing that the data judges.
Gully Sales works across marketing, sales and revenue operations, so we know how these numbers get used, and when a channel is being blamed for what is really a follow-up problem.
The rules outlast the people.
Taxonomy, credit rules and decision thresholds are documented, so a new marketing executive inherits a method instead of a spreadsheet that nobody left in the company can explain.
We do not sell you a platform you cannot maintain.
Where your existing CRM and analytics can carry the work, we use them. New tools are proposed only when the gap is real and somebody in your team can own them after we leave.
Where it applies
The same service, in different businesses.
Industrial equipment manufacturing
- The situation:
- Enquiries arrive from a listing site, from exhibitions and from dealer references, and the sales team quotes for months before an order is confirmed.
- How it applies:
- A cohort view follows each month's enquiries through to order, with first touch credited to discovery and the self-reported answer used as a check on the offline routes.
- Likely benefit:
- Exhibition spend is judged on the orders it eventually produced, not on the enquiry count taken during the week of the show.
Multi-clinic healthcare
- The situation:
- Most patients call rather than fill a form, and the person at reception records nothing about how they heard of the clinic.
- How it applies:
- A separate tracking number for each campaign, a two-question script at reception, and a source field on the patient record.
- Likely benefit:
- You learn which locality campaign fills which clinic, instead of crediting everything to the map listing.
Building materials and hardware distribution
- The situation:
- Counter sales, dealer orders and website enquiries are all demand, but the marketing budget is defended on website numbers alone.
- How it applies:
- Offline source values added to counter and dealer entry, joined to the same taxonomy the digital campaigns use.
- Likely benefit:
- The catalogue, the trade campaign and the search spend can be compared on one sheet.
B2B professional services
- The situation:
- Referrals and repeat clients bring most of the revenue, but the marketing report shows only paid search, so every budget conversation is defensive.
- How it applies:
- Referral and existing-client sources get their own values and are reported beside paid channels, with content recorded as an assisting touch.
- Likely benefit:
- Marketing is credited for the trust-building work referrals depend on, and paid spend is sized honestly against it.
Education and training
- The situation:
- Admission enquiries spike by season, come through several social channels, and are then followed up on WhatsApp by counsellors.
- How it applies:
- Campaign tagging by intake, WhatsApp enquiries logged with a source, and admissions traced by cohort rather than by calendar month.
- Likely benefit:
- You know which channel filled which batch, and can plan the next intake's spend on that evidence.
Interior design and home services
- The situation:
- Homeowners find the studio on search and social, take weeks to decide, and many arrive through a friend who was a past client.
- How it applies:
- First touch and self-reported source reported side by side, with a past-client referral value added to the taxonomy.
- Likely benefit:
- The word-of-mouth engine becomes visible enough to invest in deliberately rather than to hope for.
Direct-to-consumer e-commerce
- The situation:
- Each platform dashboard claims the same orders, and the revenue they claim between them is more than the business actually received.
- How it applies:
- A reconciliation between platform claims and order records, with a position-based view used for repeat purchase journeys.
- Likely benefit:
- Ad budgets are set against reconciled revenue instead of against three overlapping platform reports.
Questions buyers ask
Before you enquire, the answers you will want.
Which attribution model is right for our buying cycle?
It depends on how long your customers take to decide. A short cycle with one or two touches can be run on last touch alone. A long, multi-touch cycle needs first touch for discovery and a position-based or linear view beside it, so the channel that started the journey is not erased by the one that closed it. We choose the views with your sales team before anything is built, and write down which decision each view is allowed to inform.
How long does an attribution engagement take?
It depends on how many enquiry routes you have, how much CRM correction is needed, and whether we build the tracking or specify it for your developer. The audit is short. The build runs in phases, each ending in something working: taxonomy and tracking, then the CRM connection, then reporting and decision rules. We agree the sequence at the start and do not commit to a calendar before seeing the state of your data.
What access does an attribution build need?
Access to your CRM or enquiry records, your ad and analytics accounts, your website or the developer who maintains it, and the phone numbers used in marketing. From your people we need time with the sales lead to map the journey, time with whoever maintains the pipeline, and a decision on who will own the monthly pack afterwards. Most of the effort sits with us. The part only you can do is deciding the credit rules.
How is success measured?
Against the baseline recorded during the audit, on the metrics in the measurement framework: data completeness, attribution coverage, self-reported match rate, reporting time, platform to CRM variance, acquisition cost by source, conversion movement and budget efficiency. There is one further test that matters more than any of them. Does the monthly review now end in a spend decision that names the pack as its reason?
What sits outside an attribution build?
We do not run the campaigns the model then judges; that is a separate service and kept separate on purpose. So are cleaning and enriching the historical lead database, building live dashboards beyond the report pack, and migrating your CRM. We also do not backfill source data for enquiries that were never recorded, because that cannot be done honestly. The audit will tell you which of these matter for you, and in what order to take them.
Most of our enquiries come by phone and WhatsApp. Can those be attributed?
Yes, and for many Indian businesses they carry the majority of demand. Separate tracking numbers, a short script for whoever answers, click-to-WhatsApp links that carry the campaign through, and a source field the team fills at first contact will together cover most of it. It is never perfect. That is why a self-reported question on the form and at first contact runs alongside the technical tracking as a check.
Our ad platforms report more leads than our CRM shows. Which one is right?
Usually neither on its own. Platforms count conversions using their own attribution windows, and often claim a person who would have come anyway. Your CRM counts what was actually recorded, and misses whatever was never entered. We measure the gap, explain the part of it caused by windows and the part caused by capture, and set the CRM as the number of record for spend decisions once its capture is reliable.
Do we need to buy an attribution tool?
Usually not. Most SMBs get what they need from tagging done properly, their existing CRM fields, analytics they already pay for, and a call-tracking arrangement. A dedicated tool is worth discussing only when volumes are high, journeys are genuinely long and someone in your team can own the tool after handover. We will tell you plainly if the honest answer is that your current stack is enough.
4 more questions
How is this different from campaign reporting or a marketing dashboard?
Campaign reporting tells you how each campaign performed. A dashboard displays numbers that have already been defined. Attribution decides who gets the credit in the first place, which is the layer underneath both of them. Build the dashboard first and you automate a number nobody agrees with. Once attribution settles the credit rules, reporting and dashboards become straightforward, and we can link you to those services.
Our sales cycle runs for months. Will this show anything useful soon?
The capture side improves quickly. Within the first cycle you will see which sources your enquiries arrive from, which is already more than most businesses can say. The revenue side takes as long as your cycle does, because a deal has to close before it can be credited. That is why we report enquiry cohorts through to revenue, so you can watch this month's enquiries mature rather than waiting in silence.
Can walk-ins and word of mouth ever be attributed?
Not by tracking, but they can be recorded. A source value for word of mouth, a value for past-client referral, and a habit of asking the question at first contact will give you a rough but usable share. It is self-reported, so treat it as directional rather than exact. The value is in seeing how large that share is, because referral-heavy businesses often overspend on paid channels to compensate for it.
What happens if the data says our favourite channel is not working?
That is exactly why the decision rules are agreed in advance. Before the first report we set the minimum volume before a source may be judged, the period it is judged over, and the threshold at which spend changes. If a channel falls short, we look first at whether it is a discovery channel being judged on last touch, then at follow-up quality, and only then at the spend itself.
Talk to us
Request a marketing operations assessment before the next budget is set.
You get an opinion before you get a scope. We look at a sample of your recent enquiries, see how far each one can be traced, and tell you honestly how much of your demand is traceable at all.
- No obligation and no sales script
- A reply from someone who does the work
- Your details are never sold or shared