Sales Audit for a Digital Agency: Where Leads and Money Get Lost

In this case, the sales audit of a digital agency revealed something unexpected. Demand was there; the problem was sitting inside. Sales relied on the experience of the director (who is also the owner), while the team worked without clear qualification, consistent CRM discipline, an evidence-based presentation, or stable follow-up rules.

Next, let’s tackle the main question. Where exactly was the agency losing potential sales? At the entry point, in qualification, between contacts, during the presentation, or already at the level of production capacity? We won’t name the brand because of how detailed the description is and the signed NDA, but the numbers are real, and you’ll probably recognize some of them in your own business.

A digital agency came to us that specializes in PPC advertising, running ad campaigns, and managing the performance side (that is, it works on generating leads for its clients). Leads come from paid advertising (the main channel is Meta Ads), organic search, partners, and client referrals. Google Ads hasn’t worked as a source of clients for the agency itself in recent months.

The process looks like this. An SDR takes inbound requests, does the initial qualification, and sets up a meeting. Sales runs the first meeting, collects the brief, and gets access to the ad accounts. Then the PPC team spends 3–5 business days preparing a free diagnostic and strategy. At the second meeting, they present the diagnostic and the offer, issue an invoice, and take 100% prepayment for the first month.

It sounds systematic and well-structured. But the key deals are closed by the director. He’s the one who presents the offer brilliantly, argues the value, and nudges the client toward a decision. Hence the hypothesis of the audit. The problem isn’t that there’s no process at all, but that it’s poorly formalized and therefore can’t be reproduced by the team. To test this, we first need to understand what to look at.

An audit doesn’t begin with the question “why are sales low.” It begins with a map of the customer journey. A lead comes in, goes through qualification, a brief is filled out, then comes the presentation, the contract, and the launch. If you look only at the sales team, you’ll see the symptom but not the cause. Some losses sit in marketing, some in production, and a separate report on managers won’t show them. That’s exactly why a full sales audit covers the entire customer journey, not just the work of managers.

That’s why two layers are checked at once. The first layer is data. The CRM shows where deals were moving and how many got stuck. The second layer is conversations. Call recordings show how a manager asks questions and proves value. Only together do these layers give a diagnosis, rather than a set of pretty percentages. In our case, we checked the following.

• CRM structure and deal progression through stages, including skipped steps and required fields.
• Handling of inbound requests, meaning response speed and the share of missed calls.
• Quality of qualification and how complete the brief is, as well as which criteria live only verbally.
• Managers’ calls, offer presentation, and use of proof points.
• Follow-up, tasks, and the next contact date for every deal.
• Lead sources, deal duration, and reasons for losses.
• The production team’s workload and the real number of available slots.

Once the journey map is built, the main thing becomes clear. It turned out to be not in the numbers, but in the people.

The director has accumulated valuable experience. He understands who a good client is, knows how to run discovery, knows PPC economics, and ties the service to the business goal. He handles objections on autopilot. It’s a strong sales technique, and it has proven to work.

The problem is that this entire technique lives in one head. Managers don’t get ready-made clear qualification criteria, a case library, a set of proof points, standardized discovery, an offer presentation script, or argumentation rules. As a result, Sales works more like an assistant who takes notes at the meeting while the director sells. And the manager still gets the full bonus for those deals, which isn’t their fault but a consequence of the process not being documented.

It’s a trap. The owner becomes a mandatory participant in key sales and, at the same time, the main constraint on growth. The more leads, the longer the queue to the director. Now let’s look at the numbers that confirm this diagnosis.

In this case, the numbers speak far more honestly than the manager’s gut feeling. Most notably, the audit revealed several weak spots at once, not just one. Conversion in the “Long Sales” funnel turned out to be below one and a half percent. Based on the CRM, only a small share of deals went through qualification. And every second call at the offer presentation ended without a contract or invoice.

There are strengths too. The median lead processing time was 11 minutes, which is a good result. Right next to it is an unpleasant fact. Four out of ten inbound calls go unanswered, even though the call load on the team is small. The agency responds quickly to the people it manages to reach and loses those who call in themselves.

Now let’s go through the costliest problems in order, starting with qualification.

The agency’s qualification criteria were as follows: a niche from the agency’s list, a budget from $2,000, an existing website, the client’s prior experience launching campaigns with contractors, and realistic expectations. The problem is that they were stated verbally at the meeting, while the relevant sheets in the funnel file were empty. The SDR worked from experience, not from a document, and that hits scaling right away.

The brief was supposed to have about 10 required items, but in practice managers consistently completed 3–5. Further down the funnel came leads about which nobody knew how well the client fit, what their budget was, what their goal was, what they were already using, who made the decision, what constraints existed, and why the project was relevant right now. Sales accepted such a lead and redid someone else’s work, and sometimes asked the client the same questions twice.

The cost is measured in days of the team’s work. Of the lost deals, 41.4% were screening-out that should have happened on the first call. These are clients with unachievable goals (22.6%), from an unsuitable niche (11.3%), or requesting a service the agency doesn’t provide (7.5%). Each such client consumed 3–5 days of the PPC team’s work on a free diagnostic. If qualification doesn’t protect against this, the question arises of what the CRM itself shows.

The most telling example here is the “Lead Qualification” stage. Over 12 months, 77 deals were recorded at it, while 335 entered the sales funnels. That means 258 deals reached sales without ever being at the qualification stage. Managers simply moved cards from “New Lead” to another funnel. A CRM system check confirmed that stages are skipped in roughly 40% of deals.

Why is this critical? When a deal can jump from a new lead straight to a proposal, bypassing qualification, brief, and discovery, the manager goes blind. He doesn’t know the real conversion between stages, doesn’t see where the loss occurred, doesn’t understand how many quality opportunities are in the pipeline, and can’t trust the forecast. The qualification rate of 18.8% reflects not the quality of selection but how often the stage is filled in.

It’s a paradox. The CRM formally exists, leads move through it, and yet you can’t manage sales with it. It’s like a navigator that shows a pretty map but doesn’t know where you are right now. And even when a deal does move, often nobody carries it any further.

This is one of the costliest problems in the case. In 80–85% of the deals checked, the rules for repeat contacts weren’t followed. A lead didn’t pick up once, and the deal stalled. The client received information, and then silence. In half of one manager’s tasks there was no specific deadline, and wording like “feedback” didn’t say what exactly needed to be achieved in the next conversation.

Add technical gaps to that. Telephony is integrated with the CRM, but leads aren’t created from it automatically. There’s no integration with messengers and email, even though clients are often moved to written correspondence. There are no automatic tasks or reminders, so people remember leads manually, and old opportunities sit in cards for months. The next contact date is entered as a formality and isn’t agreed with the client.

Now think about what’s behind this. The lead has already been attracted and paid for by marketing. The money is spent, the interest is gained, and the company doesn’t guide the person through the process. This is the most frustrating loss, because it’s the cheapest to fix. But even getting a lead to the presentation isn’t enough if there’s nothing to show there.

The call review produced harsh numbers. The presentation block scored about 13.3%, while the use of cases and specific numbers in conversations was 0%. Managers explain well enough what the agency does, what services it provides, and what it will do. But that’s not enough for the client.

The PPC market is saturated. Almost every client has already tried advertising on their own or with a contractor and is comparing several agencies at once. The same things are heard on calls. “You’re the same kind of agency as everyone else.” “I don’t see your strong side.” “They showed both cases and numbers.” Clients say outright what they’re missing, and it’s not a discount at all. They need proof, predictability, clear economics, differentiation from others, and a mechanism for reducing risk.

In response, managers often offer ten years of experience, a story about the team, a mention of conversions from their leads to sales for clients, a promise to test different approaches, and at the end a discount. When asked to show results, one manager cited confidentiality without offering a single case instead. A service can’t be touched before buying, so without proof it turns into “just another PPC agency.” And here the director steps onto the stage again.

In this story, the director gets involved in offer presentations, argumentation, key decisions, and complex sales. He effectively makes the decision on every lead because “he has more experience.” The sales manager himself says he “mostly takes notes,” and the director admits he hasn’t had time to teach him.

A leader’s involvement in deals isn’t a problem in itself. It’s good when an owner feels the market and gets involved in important sales. The trouble starts when, without him, the quality of the sale drops noticeably. Then he becomes the bottleneck all deals have to pass through, and he still has only 24 hours in a day.

This dependence hits growth in several places at once. The number of parallel deals is limited by the director’s calendar. Scaling the team is slow because a newcomer can’t rely on a documented technique. Managers have no autonomy, and hiring one more Sales rep “without changing the processes” is impossible. But even when the director is ready to step in, the offer first has to be prepared.

According to the audit data, it took about 10 days on average from the brief to the offer presentation. The average cycle of a won deal was about 14 days. Simple arithmetic says that most of the time goes not on communicating with the client but on the company’s internal preparation.

The process also offers an explanation. By regulation, the PPC team does the diagnostic in 3–5 business days, but, according to the manager, it picks up the work on the last day of the deadline. On top of that, the diagnostic doesn’t start until full access to all accounts is received, even though part of it could be analyzed earlier. Meanwhile, the client compares several agencies and may receive a competitor’s proposal first.

It’s important not to jump to the conclusion that 10 days is too long for the whole market. It’s a problem only in the context of this particular company’s cycle. You need to check who prepares the offer, how many people are involved in the process, where the queue forms, which data is collected manually, whether calculations can be standardized, and whether every lead needs equally deep presales. For category A clients, speeding up can deliver an advantage right away. But a funnel where time has stopped being counted at all looks quite different.

The numbers here are simple. 707 leads over all time, 10 sales, and a conversion of about 1.4%. It looks like a verdict on the managers, but there’s no need to rush to conclusions. The funnel contains deals going back to 2022–2023, it’s worked “when there’s free time,” and the director himself is planning to abandon it. It’s more of a warehouse than a working tool.

Before criticizing the sellers’ skills, you need to check which leads ended up there and how well they were qualified. How many of them are old and dormant deals, what sources they came from, whether there were real contacts, and how many potential deals are actually alive. One detail is already telling. Of 23 lost deals, almost half (47.8%) were lost because the client had no website. That’s a basic criterion, known as early as the first call.

Low conversion may be a consequence of poor CRM architecture and funnel quality, not just weak sales. Moreover, the database has potential. Over a year and a half, some clients may have built a website, and a one-time reactivation with a single question about it could bring deals without a single new lead. What remains is to understand where leads come from and what they cost.

According to the report, partners delivered about 54% conversion, referrals about 70%, and Meta about 33%. In other sections of the audit, the numbers were slightly different. There, Meta in previous periods delivered 25–43%, the average for the year was about 40%, and for partners and C-level contacts it reached 80–100%. Periods and samples in different parts of the report didn’t match.

That’s why the numbers need to be reconciled before publishing final conclusions. You can’t turn them into the thesis “partners are better than Meta.” A warm partner lead and a cold ad request are simply not comparable in terms of warmth. On top of that, a manager’s result depends on which leads they were given. If warm leads go to one employee and cold ones to another, the final statistics deceive.

The right conclusion is different. Channels should be compared on a single sample, over one period, and with the same definition of conversion. Read more about how to do this in our article on analyzing the effectiveness of sales channels. Warm leads should be distributed evenly among managers so that each can be evaluated objectively. Such care is worth the effort, especially when the next problem isn’t about sales at all.

This block shows the link between sales and delivery (simply put, between the sales department and the production department). In August, the company planned about 12 slots, closed 4, and about 6 more launches were pushed to September. Not because they couldn’t sell, but because the team physically couldn’t take on new clients. The sales department worked in a mode of artificial braking, and the plan-vs-actual didn’t reflect the real commercial opportunity.

The logic is simple. If the sales team speeds up and the PPC team doesn’t accept clients, a whole trail of problems appears. The start is delayed, client dissatisfaction grows, revenue gets postponed, pressure on delivery increases, and quality may drop. A sale without the ability to launch isn’t a sale but a promise with a delay.

A direct rule follows from this. The sales plan must account for the real capacity of the production team. Setting a goal of 12 slots when the PPC team isn’t taking new projects until mid-month means building failure into the plan itself. Now let’s put all the loss points we found on one map.

If you set the individual problems aside and look at the whole funnel, the picture is coherent. Losses happen not in one place but at every joint of the process. Money leaks out at points where one person hands a lead to another or where the process relies on personal memory. Entry, qualification, touchpoints, presentation, and production each contribute their share of the overall leakage.

It’s important that many losses are interconnected. Weak qualification overloads the PPC team. An overloaded team delays the offer. A long wait for the offer gives a competitor time. And when there are no cases at the presentation, the agency’s chances shrink even more. One hole drags another along with it.

We have the map, so it’s time to decide where to start the treatment.

The first package of changes looks boring, but it’s the foundation for everything else. You need to tidy up the CRM stages and describe entry and exit criteria for each. Introduce required fields, standardize qualification, and set the rule that every deal has a next step with a date. Set up reminders, define lost reasons, and clean out old deals that have been sitting there since 2022.

Why is this the first priority? Because without proper data, you won’t understand whether the other changes helped. If today’s “qualified” leads reflect only how often the stage is filled in, then any new metric will measure noise. Order first, optimization second.

Separately, it’s worth prohibiting stage skipping. Then conversions between steps become honest, and the manager can look weekly at exactly where the funnel breaks, by manager and by source. That’s no longer a month-end report but a working tool. But order in the CRM won’t sell the service on its own; it needs to be backed up with arguments.

The next step gives managers what they lacked on every call. The audit recommends preparing 5–7 strong cases with numbers, taking NDAs into account, a short company presentation, a formulated positioning, and a USP (unique selling proposition). Standard scenarios and argumentation for different client pain points are also needed. Cases are best presented visually and tailored to specific niches.

There’s one main test for any material. It should help the manager answer the client’s question: “why should we hand our ad budget to you specifically?” If the answer sounds like “we have ten years of experience,” it doesn’t work. If it sounds like “here’s a similar task, here’s the result, and here’s the logic by which we got there,” the client hears proof.

There’s a nice bonus too. When a manager has a clear argument, they stop slipping into discounts and defensive answers. At the same time, an objection matrix appears for phrases like “expensive,” “I’m comparing with others,” and “give me a guarantee.” The proof is ready; now we need to divide who is responsible for what.

Roles at the agency are blurred, so first they need to be separated. The SDR is responsible for first contact, qualification, collecting required data, scheduling discovery, and handing off the opportunity. Their task isn’t to sell the service but to pass on a ready, clear lead. Sales takes on discovery, the brief, the diagnostic, the presentation, the offer, objections, negotiations, and the contract.

Right now, scripts help collect the brief for the PPC team but barely help sell the service. The call looks more like a questionnaire than a dialogue. The script has no questions about start timing, budget, previous results, criteria for choosing an agency, or the decision-maker. There’s no scenario for a poor-quality lead, and the handoff from SDR to Sales isn’t regulated anywhere.

For each stage, you need to define questions, a goal, a criterion for successful completion, and the next step. Then the manager knows not only what to say but also when the stage can be considered closed. Separate scripts, a mini-regulation for lead handoff, and a message bank for typical situations come together into a single system. But there’s no need to write it from scratch, because the necessary material already exists.

The most valuable part of the audit is that nothing has to be invented. The sales technique already works; it’s just not written down. So the first action is very simple. Turn on recording for discovery, briefings, offer presentations, and negotiations. Right now, such meetings aren’t recorded because there was no requirement, not all clients agree, and people forget about the button.

A working set is assembled from the recordings. A playbook describing the sales steps, a library of questions, value formulations, answers to doubts, case examples, and criteria for a good deal. In essence, it’s packaging the manager’s expertise into processes, scripts, checklists, and training materials.

The payoff is twofold. A new manager gets up to speed faster, and the director stops being a mandatory participant in every conversation. In addition, recordings make it possible to assess how Sales performs at key stages, which is impossible right now. It’s the most direct way to turn “experience in someone’s head” into a system. Next comes the quickest improvement on the list, and it’s related to the phone.

What’s needed here is an SLA, meaning clear response times. You need to define the time of first response, what to do about a missed call or a failed connection, when to write after a meeting, when after the offer, and what to do with an inactive opportunity. For each case, the number of repeat contacts, intervals, and channels are set. Conditions are also needed under which a lead goes to reactivation or is closed as lost.

This is one of the quickest improvement points. 41% of inbound calls go unanswered, while the overall processing speed is good, with a median of 11 minutes. For won leads, first contact happened within 15 minutes. So the system knows how to react quickly; it just lacks discipline at the joints. Missed calls can be fixed without a budget and without hiring.

The company has already paid to acquire the lead. If discipline is restored, the effect will appear without increasing the ad budget. This is exactly the situation where you can get a result simply by no longer losing what you’ve already gained. But one channel deserves a separate conversation, because right now it isn’t touched at all.

The partner channel is the strongest by conversion, and yet it belongs to no one. A rate of 10% for a cold lead and 5% for a loyal one looks reasonable, because closing a partner lead is objectively easier. The problem is elsewhere. The incentive covers closing the lead but not generating it. Nobody is responsible for growing the number of partners and for regular referrals from clients. The owner says outright that the partner program is “not being pushed.”

Several steps were proposed for the case. Form a partner ICP, meaning a profile of a suitable partner, define their benefit, and assign a channel owner. Then create a separate partner pipeline and track every lead passed on. Measure the qualified rate and win rate and regularly keep in touch with those already in the database. On the website, the terms for partners, according to the audit, are already published, so the foundation is there.

One caveat remains. Before declaring the channel the best by performance, the audit’s source numbers need to be double-checked, as we discussed above. But even after verification, this direction doesn’t put load on the PPC team, which is already at its limit. Now that the changes are defined, we need to understand how to measure the result.

After the CRM is normalized, metrics stop being fuzzy and can be calculated every week. There’s no need to build a complex dashboard of twenty widgets. A set that answers the questions “where does the funnel break” and “who is failing to close the deal, and why” is enough. The agency simply didn’t see many of the metrics in this set before.

Next, it’s worth distinguishing three groups. Speed and discipline metrics, funnel metrics, and resource metrics. Each group answers its own question, and together they give a picture you can manage by, rather than guess. Here’s what to keep under control.

• New leads, share processed, response speed, and missed inbound calls.
• Qualified rate and discovery rate, meaning how many leads pass selection and reach a deep conversation.
• Conversions by funnel stage, from offer and presentation to contract, plus overall win rate.
• Deal cycle and offer preparation time, to see where a queue builds up.
• Deals without a next step and loss reasons, which reveal gaps in the process.
• Conversion by source and by manager, compared on the same sample.
• Production capacity, so the sales plan doesn’t diverge from the real ability to launch.

The order matters. First data quality, and only then automated dashboards, otherwise beautiful charts will paint a beautiful lie. And how this affects day-to-day work, we’ll see through a “before” and “after” comparison.

Right now the process looks like this. A lead goes to a manager, who acts on their own experience. The brief is filled in partially, the offer takes long to prepare, and the director helps close the deal. Some leads get lost along the way, and nobody really knows where exactly.

After the changes, the picture is different. A lead comes in and immediately falls under the SLA, meaning it’s responded to within the agreed time. Then it goes through qualification against written criteria, after which the brief is filled in. Next come discovery, a showing of relevant cases, a standardized offer, and a presentation. After it, a mandatory follow-up, then the contract and client launch, taking the PPC team’s real capacity into account.

The main change is that dependence on individuals is replaced by a managed process. The director can still get involved in complex deals, but he’s no longer the only carrier of the technique. The manager understands what to do at each stage, and the head sees where the process breaks. What remains is to ask how much this applies to other agencies.

The case doesn’t mean every agency has the same problems. But it gives a set of signs by which you can tell whether a similar story lives in your company. If you recognize at least three items, it’s worth conducting a sales audit of your digital agency before you have to increase your ad budget. To understand when a sales audit is needed, the signals below will help.

The signals worth paying attention to look like this.

• The owner closes the best deals, while Sales without the expert sells noticeably worse.
• Commercial proposals take long to prepare, and the client has time to get competitors’ prices.
• The CRM looks filled in, but the numbers in it can’t be trusted because stages are skipped.
• The funnel has many “hanging” leads without a next contact date.
• There’s no clear qualification, and some of the rejections could have been predicted on the first call.
• Cases sit on the website but aren’t used in conversations with clients.
• Marketing brings in leads, while production is already overloaded and can’t launch projects in time.

These are diagnostic signs, not a verdict. Sometimes fixing one joint is enough to get things moving. Now let’s sum up.

The audit showed that even an agency with working acquisition channels and strong expertise loses a significant part of its commercial potential if sales aren’t built into a reproducible system. Leads vanished between touchpoints, deals moved past qualification, and managers lacked proof for the presentation. At the same time, the company did have its own sales technique; it just lived in the director’s head. The goal of an audit isn’t to find “bad conversion.” It’s needed to understand at which stage the company stops managing the deal and what exactly in the process needs to change.