In short (TL;DR): In 8 out of 10 client inquiries we hear: “What guarantees do you give?” The honest answer: none. Not because we lack confidence, but because guaranteeing a specific ROAS or number of leads is either manipulation or ignorance of the economics of paid advertising. Good agencies don’t give guarantees — they give something more important: transparency of process. Honest forecasts as a range, responsibility for the entire funnel, and a partnership with shared risk. In this article we break down why guarantees in marketing don’t work and what to offer instead.
The symptom: “Guarantee the result or we won’t work with you”
The client arrives after 2-3 failed launches with other agencies. The budget is burned, there are no results, trust in marketing is at zero. The logical reaction: “I want guarantees. Guarantee me a ROAS of 4, or 50 leads a month, or give the money back.”
It sounds reasonable: if the agency is confident, why not take responsibility for the result? The problem is that this request confuses two concepts: the agency’s responsibility and control over the factors that affect the result. In our practice roughly 70% of clients who ask for guarantees are actually asking for protection from a repeat of past negative experience. The other 30% simply don’t understand how paid advertising works and what a marketer’s real responsibility is.
When an agency agrees to guarantee a number of leads or a specific ROAS, it’s a red flag. Either they set the bar so low it’s impossible to miss (and you overpay). Or they manipulate the metrics (showing a “technical ROAS” that ignores all the costs). Or they simply take on the risk and then vanish when it doesn’t work out.
What ordinary agencies do: guarantees as a marketing trick
There are several ways agencies “give guarantees” without bearing any real risk:
- A guarantee on a technical metric. “We guarantee a 2% CTR” or “CPC no higher than $0.50.” Sounds specific, but these numbers have no direct link to the client’s profit. You can have a 5% CTR and still burn the budget on irrelevant traffic. One of our clients’ previous agency showed a 3.2% CTR and was proud of it, but the click-to-lead conversion was 0.8% instead of the market average of 2-3%. The guarantee was met, and the business was losing money.
- ROAS that ignores all the costs. “We guarantee a ROAS of 3.” But the calculation excludes: the agency fee (usually 15-25% of the budget), the cost of producing creatives, the manager’s time processing leads, the discounts and promotions given to lift conversion. The result: a ROAS of 3 on paper, and a blended CAC that puts the business in the red.
- A guarantee on an absolute number. “We’ll bring 100 leads a month.” The agency just pours in traffic until it hits the number, regardless of quality. For one e-commerce client the previous agency brought 847 leads in a month (the guarantee was 500). The problem: 62% of them never placed an order, and another 15% returned the goods. Real clients came to 195, while CAC grew 140% because of chaotic scaling.
- Free work until results. “You don’t pay until you get your first sales.” This only works in one case: if the agency takes clients with the simplest possible funnel and high conversion (i.e. everything already works and just needs traffic). Complex niches, long sales cycles, B2B — you can’t go there with such a guarantee, because payback may take 6-12 months.
The result of these approaches is always the same: the client gets a fulfilled “guarantee” on paper but no profit. And the agency absolves itself: “We met the contract terms, here’s the report.” In 3-6 months the cooperation ends, the client goes looking for the next agency and asks for guarantees again.
The real reason: why guaranteeing a result is impossible
The problem isn’t that marketers don’t want to take responsibility. The problem is that the result depends on factors the agency does NOT fully control. Here are the real factors that make guarantees in marketing impossible:
1. The platform auction is a black box. Google Ads and Meta Ads run on an auction model. The cost per click depends on dozens of competitors, their bids, the quality of their ads, seasonality, macroeconomics. We can optimize campaigns, but we can’t control how many competitors show up in the auction tomorrow or what budgets they have. In December 2023 one of our clients (online education) saw CPC rise 38% simply because two big players launched New Year campaigns with aggressive bids. We anticipated this and built it into the forecast, but guaranteeing a stable CPC is impossible.
2. Conversion depends on the whole funnel, not just the traffic. An agency can bring quality traffic. But if the chat on the site doesn’t work, the manager replies after 4 hours, or the offer isn’t competitive, conversion will drop. We’ve seen situations where CPL was perfect ($12 against a market of $18-22). But lead-to-sale conversion was 3% instead of 15-20%, because the sales team wasn’t handling leads by the script. Who’s to blame? The agency brought the lead, and the sale didn’t happen.
3. Testing is always a risk. Every launch starts with hypotheses: which audiences, which creatives, which offers will work. The first 2-4 weeks are the testing phase, when we gather data. Some hypotheses don’t work — that’s normal. But if an agency has guaranteed a result from the first month, it’s either lying or it won’t test anything new (and you’ll lose the chance to find far more effective combinations).
4. Macroeconomics and seasonality. In February 2022 a great many businesses in Ukraine lost 60-90% of sales because of the war. No agency could have predicted that. Less dramatic examples: seasonal demand (New Year, Black Friday), changes in purchasing power, a new competitor entering the market. All of this affects the result and isn’t controlled by the marketer.
5. Changes to the product/offer on the client’s side. The client can change the price, delivery terms, the range, the landing page — and that instantly affects conversion. We had a case where a client raised prices 20% without warning (because of rising costs), and conversion fell 35%. We saw it in the dashboard and reacted quickly (changed the offer, added installments), but if there had been a guarantee on the number of sales, the agency would formally have failed to meet the terms, even though the cause had nothing to do with traffic.
Here’s a comparison of what the client assumes versus what actually affects the result:
| What the client assumed | In reality (the real factors) | Agency’s share of influence |
|---|---|---|
| Result = quality of campaign setup | Traffic quality + funnel conversion + lead handling + product | 30-40% |
| ROAS depends on the agency | ROAS = (traffic × conversion × AOV × repeat rate) / (ad spend + fee + creatives) | 25-35% |
| The number of leads can be guaranteed | Number = budget / CPC / landing page conversion / competitors’ auction | 20-30% |
| A good agency = a stable result every month | Results fluctuate ±15-25% due to seasonality, testing, market changes | 40-50% |
As you can see, the agency controls 20-50% of the factors that affect the final result. The rest is the client’s business economics, external factors, the sales funnel. Guaranteeing a result in such a system is like guaranteeing the weather.
Our approach: what to give instead of guarantees
If guarantees don’t work, what gives the client confidence? At LeadPrice we give five things that are far more valuable than a formal guarantee on paper:
1. Diagnosis before launch — honest forecasts as a range
Before taking on a client, we run a free audit of the entire economics: unit economics (CAC, LTV, margin), the funnel across 5 stages (click → lead → qualification → sale → repeat), the competitive environment in the auction. Based on this we give a forecast not in the format “we guarantee 100 leads” but as a range:
- Pessimistic scenario (if the hypotheses don’t work and the auction is expensive): CPL $25-30, sales conversion 8-10%, blended CAC $280, payback in 4-5 months
- Realistic scenario (the baseline we aim for): CPL $18-22, sales conversion 12-15%, blended CAC $160, payback in 2.5-3 months
We DON’T give an “optimistic” scenario, because it’s usually a fiction that creates unrealistic expectations. If the realistic forecast shows the business won’t pay back even in 6 months, we honestly say “no” and don’t take the project. Roughly 2 out of 10 clients who approach us are turned down for exactly this reason.
2. Transparency of process — access and a real-time dashboard
All the tools are set up on the client’s accounts (Google Ads, Meta Ads, analytics). The client has full access and sees all the data in real time. We also give access to our internal dashboard, which shows the whole funnel: how much was spent → how many clicks → how many leads → how many sales → what the blended CAC is → what the LTV is → when payback comes.
No black boxes. No “trust us, everything’s fine, here’s a pretty PDF report.” The client sees the same picture we do. If something goes wrong, we discuss it at the weekly syncs and adjust the strategy.
3. Responsibility for the whole funnel, not just the traffic
We don’t say “we brought the traffic, the rest isn’t our problem.” Our responsibility is the entire cycle from click to repeat sale:
- Proposition and offer. If conversion is low, we don’t just tweak bids — we rewrite the offer, test new mechanics (installments, bonuses, money-back guarantees).
- Landing page. A/B tests of headlines, CTAs, forms. If needed, we bring in a CRO specialist (we have one on the team).
- Lead handling. If lead-to-sale conversion is under 10%, we audit the sales team, write scripts, set up the CRM.
- Repeat sales. If LTV is low, we build an email/push strategy for retention, test upsell/cross-sell.
This doesn’t mean we do everything ourselves (there are things the client has to do on their side), but we take responsibility for diagnosing problems at every stage and give concrete, prioritized recommendations.
4. A partnership model: we’re in the same boat
Instead of the “client-contractor” model (where the client pays a fixed fee and the agency turns the knobs), we offer a partnership. There are two options:
- A base fee + performance bonus model. For example: a base fee of 20% of ad spend + a 5-10% bonus if the target CAC and ROAS are reached. This aligns interests: the agency is motivated not just to spend the budget but to achieve a real result.
- A model with partial payment from profit. For long-term clients (6+ months) we can move to a model where part of the agency’s compensation is a % of the incremental profit we generate. That is, if we increased the business’s profit by $50K in a month compared to the baseline, we receive, for example, 15% of that $50K.
This isn’t a guarantee in the classic sense, but it is shared risk: we earn more only if the client earns more.
5. A clear 3-6 month roadmap with checkpoints
We don’t say “let’s start and see.” After the diagnosis we provide a roadmap with concrete actions by month and the metrics we expect to see at each checkpoint. Here’s an example for an online school client (B2C, average order $450):
| Stage | Timing | Actions | Expected metrics |
|---|---|---|---|
| Phase 1: Testing | M1 (weeks 1-4) | Launch 4 audience hypotheses, 8 creatives, A/B test of the landing page | CPL $25-35, conversion 1.5-2%, ad spend $8-10K |
| Phase 2: Optimization | M2 (weeks 5-8) | Cut ineffective combinations, scale the winners, optimize the offer | CPL $18-24, conversion 2.5-3.5%, ad spend $15-18K |
| Phase 3: Scaling | M3 (weeks 9-12) | Increase budget 40-60%, launch lookalikes, retargeting | CPL $20-26, conversion 3-4%, ad spend $25-30K, blended CAC $180-220 |
| Phase 4: Retention | M4-M6 | Email automation, upsell, referral program | Repeat rate 15-20%, LTV $680-750, payback 2.5-3 months |
At the end of every month we check: did we hit the checkpoints or not. If not, we work out why and adjust the plan. This isn’t a guarantee of results, but it is a guarantee of process: we won’t just sit and watch the budget burn.
Case: how we gave a forecast instead of a guarantee — and delivered on it
The client came after two agencies that had both promised a “guaranteed ROAS of 5.” The first agency showed a ROAS of 4.8 in the first month, but on inspection it turned out they’d counted only ad spend, without their fee (20%) or the cost of creatives ($2.5K). The real blended ROAS was 2.9, which for their margin (40%) meant working at zero. The second agency simply disappeared 6 weeks after failing to hit what it had promised.
When the client approached us, we ran a diagnosis and saw the problem: LTV was understated ($320 against a possible $550-600), because there was no retention strategy. CAC was $180-200, so paying back from the first sale alone was unrealistic.
Our forecast: pessimistic CAC $220, realistic $180-190, provided we raise LTV to $500+ through email automation and upsells. Payback under the realistic scenario — 3-3.5 months. We gave no guarantees, but proposed a fee model: a base of 18% + a 7% bonus if we achieved a blended CAC under $200 and an LTV above $480.
The result after the first 4 months: CPL fell from $28 to $19. Lead-to-sale conversion grew from 11% to 18% (we rewrote the managers’ script and added a chatbot for fast qualification), blended CAC $185, LTV $520 (we launched 6 email sequences and a cross-sell strategy). Payback — 2.8 months, better than the realistic forecast. The client extended the cooperation for the next 12 months.
FAQ: guarantees in marketing
Why do some agencies give guarantees and you don’t?
Because either their guarantees are built on manipulation (they guarantee metrics that aren’t tied to profit), or they only take very simple clients where everything already works and just needs budget. We work with complex cases: long sales cycles, B2B, highly competitive niches. Guaranteeing a specific result there means deceiving either yourself or the client. We give honest forecasts as a range and take responsibility for the process, not for a number we can’t fully control.
If you don’t give guarantees, how can I be sure you won’t burn the budget?
First, all the tools are set up on your accounts, and you have full access to all the data in real time. You see every dollar, where it went and what it delivered. Second, we provide a roadmap with checkpoints — if at the end of the first month the metrics are far from the forecast, we stop and work out why. Third, our business model is built on long-term relationships (the average cooperation is 18+ months), not on taking a client for 2 months and vanishing. Burning a budget means killing our reputation — that’s not in our interest.
What happens if your forecast doesn’t come true?
It depends on the cause. If we were wrong in our hypotheses (for example, the audience turned out more expensive than expected), we see it within the first 2-3 weeks and adjust the strategy at no extra cost to you. If the problem is on the client’s side (low lead-to-sale conversion, a weak offer), we diagnose it and give a concrete plan for what needs to change. We have a rule: if after the first month the real metrics are 30%+ worse than the pessimistic forecast and we see no way to fix it, we propose stopping the project and refund the second month’s fee. That has happened twice in the last 2 years.
What about payment for results (performance fee)?
We offer a hybrid model: a base fee (covers our work) + a bonus for hitting target metrics (CAC, ROAS, LTV). We don’t do a pure performance model (0% base fee, only a % of sales), because it creates a conflict of interest: the agency will chase volume at the expense of quality instead of building a long-term strategy. But if you’re ready for long-term cooperation (6+ months), we can discuss a model where part of our compensation is tied to the incremental profit we generate.
Can you guarantee at least a minimum number of leads?
No, because the number of leads depends on factors we don’t fully control: the auction cost, landing page conversion, seasonality. We can give a forecast as a range (for example, “we expect 80-120 leads on a $10K budget”). But guaranteeing a specific figure means promising something that depends on Google/Meta algorithms and your competitors. Instead, we guarantee the process: weekly syncs, a transparent dashboard, fast reaction to changes. And if something goes off plan — an honest conversation and a course correction.
What if I want to try working with you but I’m not ready for a long-term contract?
We work without long-term commitments. The standard model: the first month is a trial (can be terminated at any time with 2 weeks’ notice), then a monthly contract. The minimum period to see a real result is 3 months (the first month is testing, the second is optimization, the third is the start of scaling). But if after the first month you see the approach isn’t a fit, you aren’t tied in. Our goal is for clients to stay with us because they see results, not because they signed a one-year contract.
Conclusion: guarantees don’t replace trust, but trust is built on transparency
Honest marketers don’t give guarantees on a specific ROAS or number of leads, because that’s either manipulation or ignorance of the economics of paid advertising. Instead they give something more valuable: honest forecasts as a range, full transparency of process, responsibility for the whole sales funnel, and a partnership model where the risks are shared.
If you want to work with an agency that won’t promise the impossible but will take real responsibility for your result, we’re ready to talk. Order a free audit — we’ll go through your economics, give an honest forecast (even if you won’t like it) and show what our approach looks like in practice. It costs nothing, and you’ll come away understanding what really affects your results in paid advertising.
Take a look at our cases to see how we work with real clients. Or browse our services to understand what comprehensive marketing support includes.