How to choose an agency.

Where choosing an agency begins

Choosing a contractor begins not with a search but with formulating the task. As long as the task sounds like “we want it to look nice” or “we need marketing”, there is nothing to compare proposals against: every agency will slot into that frame whatever it knows how to sell.

The goal must be verifiable. “Increase the number of enquiries from the website”, “reduce the cost per enquiry”, “bring a new product to market”, “make the company findable by its name” — these are different tasks, and different teams solve them. An agency strong in contextual advertising is not necessarily strong in branding, and vice versa.

It helps to answer three questions for yourself in advance: what should change in the numbers, within what timeframe, and what budget you are prepared to allocate to it per quarter. If there are no answers, the first conversation with any agency should start with an audit, not with a brief for work.

Separately, it is worth understanding what you expect from the contractor: a pair of hands or a partner who will argue with your hypotheses. These are different formats of cooperation and different prices.

How to draw up a shortlist

Next comes market research. The list should be compiled from several sources at once, not from the first page of search results: recommendations from fellow entrepreneurs in adjacent niches, industry rankings, agencies' publications, professional communities.

First and foremost, it makes sense to look at specialisation. There are no universal agencies: every team has an area where it is strongest and areas it covers “for good measure”. If your task is promoting a service with a long sales cycle, experience with online stores will give you little.

A good sign is when the portfolio includes projects of different scale, not only big brands. It means the team can work not only with large budgets but with limited ones too.

It is sensible to limit the shortlist to three to five companies. More, and the comparison turns into endless correspondence; fewer, and there is nothing to compare proposals against. We wrote about which channels suit your business at all and what is worth entrusting to a contractor in a separate piece on choosing promotion channels.

What to look for in a portfolio

The portfolio is the most overrated part of a presentation. Logos of well-known clients say nothing: the company could have ordered a single banner and left. What works is something else — a case study with numbers.

A proper case study has a starting point, a task, what exactly was done, how long it took and what came of it in measurable terms. If the description contains only “raised awareness” and “made a modern design”, you are looking not at a case study but at a slide.

Separately, check whether the projects in the portfolio are still alive. A website built three years ago and running ever since is a more convincing argument than a pretty picture in a presentation. Our work, with descriptions of tasks and results, is collected in the case studies section.

And be careful with promises at the outset. The numbers in a case study relate to someone else's business, someone else's niche and someone else's budget. They cannot be transferred to your project as a forecast — only as confirmation that the team knows how to see the work through to a result.

Reviews and references: how to check

Insufficient vetting of a contractor is one of the most common and most expensive mistakes. Reviews on an agency's website are promotional material by definition, so look for them where the agency does not control publication: in industry rankings, in professional communities, in a search for the company's name.

The most useful format is a direct conversation. Ask for the contacts of two or three clients, ideally not only current but also former ones. The questions that actually yield something: were deadlines met, how were disputes resolved, did the team on the project change, what did not work out.

The absence of negative reviews is not a virtue. Any team with a history has projects that did not go well. What matters is how the agency talks about them: does it admit the mistake and explain what it changed in its work afterwards.

How the first conversation goes

The first meeting gives more information than a portfolio. Pay attention to who talks more: if the agency spends the whole hour talking about itself and never once asks about your funnel, average order value, seasonality and the customer's path to a deal, it will carry on the same way.

A good sign is uncomfortable counter-questions. What does customer acquisition cost now, who handles enquiries, how long does it take from enquiry to sale, where are enquiries lost. An agency that asks these intends to be accountable for results, not for the number of posts published.

Watch the language. If the answers consist of professional jargon and after the explanation you understand no more than before, ask them to rephrase. A contractor who cannot explain in plain words will not be able to explain a report either.

One more point is speed. You should not choose an agency in a single day, but there is no sense in dragging the choice out for months either: while the choice is being made, the budget is not working. A reasonable timeframe is two to three weeks to compare proposals.

How much it costs

Prices for the same task differ several times over between agencies, and that is normal: behind the figure are different scopes of work. So what needs comparing is not the final number but what goes into it.

Clarify the composition of the team on the project, the volume of work per month, whether analytics and reporting are included in the price, whether production of materials is charged separately, who pays for advertising budgets and services. Two proposals with the same price may describe completely different scopes.

The cheapest proposal usually means one of two things: either part of the work will not be done, or a junior specialist will be on the project. Both cost more than what was saved — redoing work costs more than work done right the first time.

The opposite extreme also occurs. A high price in itself guarantees nothing if there is no clear scope of work and accountability for the metrics behind it.

Common mistakes when choosing

• Deciding after a single conversation. Haste in choosing a contractor is a typical cause of a failed project: there was no time to compare, there was nothing to evaluate.
• Focusing on price alone. A cheap solution turns into quality that has to be fixed for extra money.
• No vetting. Reviews were not read, clients were not called, case studies were not opened — they trusted the presentation.
• A vague brief. The fuzzier the task is described, the higher the chance of getting something other than what you need, and of having no formal grounds for complaint.
• Expecting marketing to solve everything. A contractor influences the flow of enquiries, but not the quality of the product or the work of the sales department.
• Silence after launch. A project without regular feedback drifts off course unnoticed by both sides.

The last point deserves particular attention: a significant share of failures is not a contractor's low skill but lost communication. Regular discussion of interim results is cheaper than a post-mortem six months later. What to do if advertising is already running but there are no enquiries is covered in our piece on advertising that brings no results.

What to put in writing before the start

Verbal agreements do not survive the first conflict. Before work begins, several things should be set down in writing.

1. The metrics by which the result is assessed, and the period over which they are assessed.
2. The scope of work per month and the report format — what it must contain.
3. Deadlines and the procedure if they slip.
4. Who owns the ad accounts, analytics counters, domain and source files.
5. The handover procedure when the cooperation ends.
6. Who makes decisions on the agency's side and on the business's side.

The point about ownership of access is skipped more often than the others, and it costs the most: ad accounts registered to the contractor leave with them at parting, together with the accumulated statistics. How to read reports and calculate return on investment is covered in detail in our article on data analysis.

Choosing an agency is not buying a service but the start of a partnership in which the contractor understands your business and you understand what you are paying for. That is exactly how we work at epic name., including on special projects and non-standard tasks, where no ready-made solution exists and the price has to be justified by substance rather than a price list. For the promises you should not believe at the outset, read our separate piece.