Digital Marketing Agency in Pune: Why Cheap Leads Are Not Good Leads, and How to Feed Sales Outcomes Back Into Your Ads
A digital marketing agency in Pune should be judged on how many enquiries turn into customers, not on how many enquiries it generates. The way to make that possible is to track every lead from the ad click to the sales outcome, record whether it became a qualified opportunity or a sale, and send that result back to Google and Meta so their systems learn which clicks produce buyers. Without that loop, ad platforms optimise for whatever you told them to count, which is usually a form submission, and a form submission costs the same whether it came from a serious buyer or a student doing homework.
This guide is for owners and sales heads of service businesses, clinics, education providers, real estate firms, manufacturers and B2B companies, where the sale happens on a call or a visit rather than at an online checkout. If your reports show a falling cost per lead while your sales team complains that the leads are rubbish, this is the problem you have, and it is fixable. We explain the mechanism, the data you need to capture, the setup on Google and Meta, the way to define lead quality honestly, and how to read the results. We do not quote benchmark costs, because they vary by industry, location and season; you should compare against your own history.
The problem: platforms optimise for what you measure
Ad platforms are very good at finding people who are likely to do the thing you ask them to optimise for. If the optimisation target is "form submitted", the system will search for people who submit forms easily. Those tend to be people with low intent: curious browsers, people comparing prices with no plan to buy, and occasionally bots or accidental clicks.
This is not a flaw in the platform; it is a consequence of an incomplete instruction. The platform cannot know that a lead who did not answer the phone is worth less than one who booked a site visit, unless you tell it.
The result is a familiar pattern:
- The agency reports a lower cost per lead this month.
- Sales says the leads are poor.
- The agency says lead volume is up and cost is down, so the campaign works.
- The business owner has no way to settle the argument, because the marketing report and the sales pipeline live in different places.
The fix is not a debate; it is a shared definition of a good lead, and a data path that carries the outcome back to the ad platform.
Step one: agree what a "good lead" means
Before anyone touches tracking, write a short definition with the sales team. It should be specific enough that two people would classify a lead the same way.
A usable lead stage ladder looks like this:
| Stage | Meaning | Example for a clinic | Example for a B2B manufacturer |
|---|---|---|---|
| New lead | A form, call or message was received | Appointment request form | Quotation request form |
| Contacted | Someone reached the person | Front desk spoke to the patient | Sales rep spoke to the buyer |
| Qualified | Meets your basic criteria | Lives within the service area, treatment is offered | Right product, real quantity, decision maker |
| Opportunity | A concrete next step is booked | Appointment confirmed | Quotation sent or site visit booked |
| Won | Became a customer | Patient attended and paid | Order placed |
| Lost / junk | Not a fit, spam or wrong number | Wrong city, job seeker | Student project, supplier pitch |
Two cautions. First, qualification criteria should be the sales team's, not the agency's; the agency can suggest, but the people who answer the phone know what a good lead looks like. Second, keep the number of stages small. Five or six that people actually update is better than twelve that nobody does.
Your CRM, or even a shared spreadsheet if you are small, needs a status field that follows this ladder. If leads currently live in WhatsApp chats and someone's memory, the first step is to give them a home. Our article on lead capture and CRM integration explains how forms feed a CRM, and our CRM service is one option if you need to build one.
Step two: capture the click identifiers with every lead
For the outcome to be tied back to an ad, the lead record must carry a trace of the ad click that produced it. The mechanism differs by platform.
Google Ads. When someone clicks a Google ad, Google can append a click identifier (known as the GCLID) to the landing page address, when auto-tagging is switched on. If your website captures that value when the visitor submits a form and stores it with the lead, you can later tell Google "this click became a customer". The value is typically stored in a form field that your developer fills in automatically from the address; the visitor does not type it and does not need to see it.
Meta (Facebook and Instagram). Meta uses its own identifiers, typically stored in browser cookies (such as the click and browser identifiers), together with the lead's contact details, which can be matched to Meta accounts when sent in the approved, privacy-conscious way. Your developer or agency should read Meta's current documentation on its Conversions API and follow it exactly, including how contact details are hashed before being sent.
Lead forms inside the platforms. If you use Meta's in-app lead forms or Google's lead form assets, the platform already has the click context, and the lead data is delivered to you through an integration or download. The same principle applies: you must record what happened to each lead afterwards and tell the platform.
Phone calls. If many of your enquiries come by phone, you need call tracking so that calls are attributed to the campaign, keyword or ad. We cover the setup in our guides on call tracking for Google Ads in India and tracking phone leads from Google Ads.
WhatsApp. Click-to-WhatsApp ads produce conversations rather than form fills. Tracking them requires the conversation to carry its source, and the outcome to be recorded in your CRM or WhatsApp tool. See our article on click-to-WhatsApp lead automation.
The practical test: pick ten recent leads and ask, for each, "which ad, keyword or campaign produced this?" If you cannot answer, you cannot feed outcomes back.
Step three: send outcomes back to the ad platforms
Once you have a click identifier on each lead and a status in your CRM, you can send events back.
On Google Ads
Google Ads supports importing offline conversions: you tell it, using the click identifier and a timestamp, that a particular click later became, for example, a qualified lead or a sale. You can upload these manually as a file, on a schedule, or through an integration from your CRM. Google also offers "enhanced conversions for leads", which uses hashed contact details from your form to improve matching when the click identifier is unavailable. Setup details change from time to time, so follow Google's current help documentation or have your agency do so, and confirm that consent requirements are met for your visitors.
The value of the import is not only reporting. Once Google has conversion data on qualified leads or sales, you can optimise campaigns toward them and compare keywords by cost per qualified lead, not by cost per form.
On Meta
Meta's Conversions API allows a business to send events from its own systems, such as "lead qualified" or "purchase", and Meta provides optimisation options based on lead quality for some accounts and objectives. Availability and names of these options change, so ask your agency to check which are available for your ad account today and to document what was configured.
Where it is available, sending quality events lets the system look for people who resemble the leads you actually valued, rather than everyone who is willing to tap a form. Our article on Meta's Advantage+ becoming the default for lead generation explains why this matters more than before, because more targeting decisions are now made by the algorithm.
Keep it honest and simple
- Send only events you actually observed. Do not send invented conversions to make a campaign look healthy. It corrupts the learning and your reports.
- Pick one or two quality events, such as "qualified lead" and "won", rather than every stage. Too many events with too few counts each give the system little to learn from.
- Allow for delay. A lead that becomes a sale in three weeks cannot report the sale in three days. Set expectations about the lag, and agree how long you will wait before judging a campaign.
- Check consent. Under India's data protection law, personal data should be collected and used with appropriate notice and consent. Make sure your forms say how contact details are used, and take advice on what is appropriate for your business. Our guide on DPDP Act website compliance is a starting point.
Step four: measure by cost per qualified lead, not cost per lead
Once stages are being recorded, the report changes shape. Instead of one number, you see a funnel by campaign:
| Campaign | Spend | Leads | Qualified | Won | Cost per lead | Cost per qualified lead |
|---|---|---|---|---|---|---|
| A | (your figure) | (your figure) | (your figure) | (your figure) | (calculated) | (calculated) |
| B | (your figure) | (your figure) | (your figure) | (your figure) | (calculated) | (calculated) |
We have left the table empty on purpose: the numbers must come from your own account, not from a generic benchmark. The point is the comparison. It is common to find that the campaign with the lowest cost per lead has the highest cost per qualified lead, and the reverse.
Some practical reading rules:
- Do not judge a campaign on fewer than a meaningful number of qualified leads. Small counts swing wildly. If a campaign has produced only two or three outcomes, treat conclusions as provisional.
- Compare like with like. Branded search, general search, Meta lead forms and click-to-WhatsApp ads attract different intent and should be assessed separately.
- Look at lead-to-qualified rate by source, which tells you about audience quality, and by landing page, which tells you whether the page sets the right expectations. A page that states prices, service area and minimum order size filters out poor-fit leads before they fill the form, which is often better than filtering them afterwards. See our article on whether ads should go to a landing page or the homepage.
- Count the sales team's response time. A good lead contacted after two days is a poor lead by the time you speak to them. If qualified rates are low across all campaigns, investigate follow-up before blaming the ads. Our article on automating sales follow-ups with a CRM covers this.
Using quality data to improve the ads
When you know which sources produce qualified leads, you can act on it in ways that would otherwise be guesswork.
Add negative keywords from junk leads. If lead notes show many job-seekers, students or "free" enquiries, add the search terms that brought them as negative keywords in Google Ads, and review the search terms report at least weekly in the early months.
Tighten the form. If low-quality leads dominate, a qualifying question such as budget range, quantity, preferred location or timeline can reduce junk without discouraging serious buyers. This trades volume for quality; decide deliberately.
Adjust geography and schedule. If leads from certain areas never convert, narrow targeting. If leads arriving after business hours go cold, change ad scheduling or add an automated first response.
Rewrite the ad. Ads that state who the service is for, and who it is not for, reduce unqualified clicks. A clinic ad that names the treatment and area outperforms, on quality, a generic "book now".
Rebalance the budget. Move spend from campaigns with high cost per qualified lead to those with lower, in steps, rather than sudden swings, and watch results for long enough to see a pattern. For budget thinking see our guides on Google Ads budgets for small businesses and Google Ads versus Meta Ads.
What to ask a digital marketing agency about lead quality
If you are hiring or reviewing an agency, these questions reveal whether they work toward sales or only toward lead counts:
- How will we define a qualified lead, and who signs it off? The answer should involve your sales team.
- How will each lead be traced to its ad, keyword and landing page? Look for specifics: click identifiers, call tracking, UTM tags.
- Will outcomes be sent back to Google and Meta? And if not, why not?
- Which account owns the data? The ad accounts, analytics property and conversion data should be in your name. Our article on what to prepare and hand over in the first two weeks covers ownership.
- What will the monthly report show beyond cost per lead? It should reach at least qualified leads and, where possible, sales. See what a monthly retainer should deliver and how to read the report.
- What happens if lead quality is poor for a month? A responsible agency has a diagnostic routine: check search terms, landing page, form, follow-up speed and targeting, rather than simply adjusting bids.
An agency that welcomes these questions is usually one that expects to be judged on outcomes.
Common mistakes
Counting every form fill as equal. Value differs; the system should reflect it.
Trusting the platform's own conversion count without reconciling to the CRM. Platform numbers and CRM numbers should be compared monthly; large gaps point to tracking faults, duplicate counting or leads that never reached the CRM.
Tracking tag on the thank-you page only. If the thank-you page can be reached without submitting a form, or the form does not redirect, the count will be wrong. Test by submitting real forms.
Forgetting phone and WhatsApp leads. If half your enquiries arrive by call or chat but only forms are tracked, the data will misrepresent which campaigns work.
Judging too early. Quality data arrives with delay; set a review point at which you judge, based on your sales cycle.
Sales not updating statuses. The loop depends on people. Make updating the CRM part of the sales routine, and keep the stage list short enough that they actually do it.
A first-month plan
If you are starting from scratch, a realistic order of work is:
Week one. Agree the lead stage definitions with sales. Audit the existing tracking: submit test forms, place test calls and confirm each appears in analytics and the CRM. Fix obvious failures.
Week two. Add click-identifier capture to the website forms and store it in the CRM. Set up call tracking if phone matters. Tag campaigns consistently with UTM parameters.
Week three. Begin updating lead statuses daily. Build a simple report joining spend, leads and qualified leads by campaign.
Week four. Set up the offline conversion import or Conversions API events for your one or two chosen outcomes, test them with a small batch, and confirm they appear in the platform.
Months two and three. Review cost per qualified lead by campaign, adjust budgets, negative keywords, forms and landing pages, and begin optimising toward quality events once enough of them exist.
None of this requires a large budget, but it does require cooperation between marketing and sales, which is why it is worth agreeing at the start.
Working with Govindani Infotech
Govindani Infotech Pvt. Ltd. is a digital marketing and web development company in Pune. Our Google Ads service and Meta Ads service are set up around the question of what a lead is worth, and our CRM work helps connect enquiries to outcomes. If you want a review of how leads are tracked from click to sale in your current setup, message our team on WhatsApp or use the contact page. We will tell you what we find, including when the problem is follow-up rather than advertising.
Frequently Asked Questions
Why are my Google and Meta leads cheap but low quality?
Because the ad platform is optimising for what it is told to count, usually a form submission, and form fills are easy to get from low-intent people. Define a qualified lead with your sales team, track it, and send that outcome back to the platform so it learns to look for similar people.
What is offline conversion tracking?
It means telling an ad platform about results that happen away from the website, such as a qualified call, a booked visit or a sale recorded in your CRM, and linking them back to the ad click that started the process. It lets you measure and optimise by sales outcomes instead of form submissions.
Do I need a CRM to track lead quality?
You need somewhere to record each lead's status, and a CRM makes that reliable. A small business can start with a shared spreadsheet, as long as every lead has a source, a status and a date. As volume grows, a CRM reduces missed follow-ups and makes outcomes easier to send back to ad platforms.
How long before I can judge whether leads are good?
It depends on your sales cycle. A lead that becomes a customer in a day can be judged quickly, while one that takes a month needs a month. Agree a review point based on your own cycle, and avoid changing campaigns on the strength of only a handful of outcomes.
Is it safe to send customer details back to Google or Meta?
It can be done in privacy-conscious ways, such as using click identifiers and sending contact details in hashed form, but you must still have a lawful basis, give visitors clear notice and follow each platform's policies. Take advice on your obligations under India's data protection law, and ask your agency to document exactly what is sent.