Best Meta ads agencies for software companies: how to judge without a ranked list

Why is every ranked list compromised?

Because of who writes them, and there are only three kinds of author.

The first is an agency writing about its own market. It appears at or near the top, the rest of the list is chosen to make that placement look reasonable, and no assessment of anyone's work has taken place. This page would be one of those if it ranked anybody.

The second is a directory or review site funded by placement fees or lead sales. The ordering follows commercial relationships. Some are transparent about it in a footer, most are not.

The third is a publisher with no stake, which sounds better until you ask what they measured. Almost always it is public information: website quality, case studies the agency wrote about itself, a founder's follower count. None of those correlate with whether your campaign works.

There is no fourth kind, because the information required to rank agencies honestly is confidential client performance data, and nobody publishes it.

Ranking agencies honestly would need confidential client performance data. Nobody has it, so every list is measuring something else and calling it quality.

What can you actually verify?

Three things, and all of them are answerable in a first conversation if you ask directly.

Who does the work day to day, by name, and how many other accounts that person runs. This is the single most predictive question and the one most often deflected. The person in the meeting is frequently not the person in the ad account, and an operator carrying too many accounts cannot give any of them attention regardless of talent.

What happens when it does not work. Ask for the last client that did not go well and what they did about it. Everyone has one. An agency that cannot produce the story is either new or not being straight with you.

What you keep when you leave: the ad account, the pixel and its history, the creative. If the answer is anything other than all of it, you are renting your own marketing asset.

None of those require trusting a list. All of them require asking.

1. Who runs the account by name, and how many others they run.
2. What happened with the last client that went badly.
3. Whether you keep the account, the pixel history and the creative.

What about case studies and reviews?

Read them, and read them for what they omit rather than what they claim.

A case study is written by the agency about its own best outcome, so the result is not the informative part. The informative parts are the starting conditions, the time it took, and what the client had to do. A case study that describes a starting point, a timeline and a client-side dependency is describing real work. One that reports a percentage increase with no baseline is describing nothing at all.

Reviews have a different failure. They are usually written at the point of maximum optimism, a few weeks into an engagement, before anybody knows whether it worked.

The stronger move is to ask for a reference from a client who left, and to ask them one question: what did you wish you had known at the start? An agency that will not connect you to a former client has told you something.

How do you tell specialisation from a landing page?

By whether they can describe your business back to you without being prompted.

Any agency can add software companies to its website. The test is whether they know the things that only come from doing the work: that your sales cycle outruns the attribution window, that a work email matches poorly against personal social profiles, that a free trial signup is a much weaker signal than it looks, that your buyer often cannot be found by job title.

If those come up unprompted, they have run these accounts. If the conversation stays on creative and audiences and could be about any industry, they have not.

Ask what they would expect to go wrong in the first ninety days with a company like yours. A specialist has a specific answer, and a generalist has a reassuring one.

Ask what they expect to go wrong in the first ninety days. A specialist gives you a specific answer; a generalist gives you a reassuring one.

How would an honest roundup be built?

It would need something no published list currently has, which is verified outcomes with the client's consent.

That means a defined metric agreed in advance, results confirmed by the client rather than the agency, disclosure of who paid for anything, and a published method for who was considered and who was excluded. It would also need to name the agencies that declined to participate, because a list of only volunteers selects for confidence rather than quality.

We are not the right people to publish that about our own market, and we would not ask you to trust it if we did.

Until somebody independent builds it, the questions above will tell you more in one conversation than any ranking will. Ask them of us too.