How to Find Affiliates for Your Brand in 10 Minutes

I spent seventeen years in the affiliate space working out of spreadsheets, guessing at domains, and firing emails into info@ boxes that nobody reads.

That is not just a confession. For most of you reading this, that is a job description you have had, or one you still have.

So here is what you are going to get in the next ten minutes of reading: the actual process for how to find affiliates for your brand, what each step costs you when you do it by hand, and what it looks like when you stop. I ran this live on the School of Chaos podcast with Matt Frary, President and COO at XPFlow, and Beth Lazar, Chief Product Officer at XPFlow. Blank account, no script, and a brand the audience picked while we were on air.

Most product webinars show you a beautiful demo that does not exist and where nothing ever breaks. It is not real. We did the opposite. Everything below actually happened, including the parts where Beth had to type in front of two hundred people.

Here is the video.

Because relationships are messy, and until recently nobody could automate messy.

Every other channel in marketing got automated years ago. Paid search and programmatic are coin operated: you put money in, you get clicks out. Affiliate never worked that way. Matt put it plainly on the episode:

"Affiliate really is relationship driven, and relationships are messy. How do you find the contacts? How do you convince someone that your link, your product is gonna make more money than the 10,000 other requests that happened today?"

That messiness is why the job drifted so far from what it was supposed to be. Matt has been doing this for thirty years, and he described his junior affiliate manager years as being a "spreadsheet jockey." He signed up for the relationship business and got reporting and manual labor instead.

The thing that changed is not that relationships got simpler. It is that AI got good at the part that was never relationship work in the first place: sifting millions of sites, structuring unstructured data, finding a working contact, and drafting a first email that references something real.

As Matt said, it is not about removing the human. It is about taking out the messy little bits so the human has time to do the part that actually requires a human.

Affiliate recruitment is the process of identifying websites, creators, and publishers whose audience already wants what you sell, finding a real person to contact at each one, and pitching them on promoting your product for a commission.

Done manually, it is seven separate jobs:

Matt described the agency version of step one: walking into the bullpen and telling the team "Robinhood, you’re recruiting for that today, get into Google search," then everyone brainstorming keywords for hours. Finance review sites. Finance app sites. Etcetera, etcetera.

That is hours of human time before a single email goes out.

Here is the whole process as we ran it live. The audience picked Robinhood, the finance app, out of the submissions. Nothing was preloaded.

That is the entire input. Beth pasted the Robinhood URL into Alfie and it read the site.

Alfie came back with keywords (one of them was "best apps for beginner investors") and a recruitment prompt built from the site itself:

"We’re looking to partner with content creators and sites that focus on personal finance, investing education, financial empowerment, and particularly those who cater to newer investors and self directed retail traders. Ideal affiliates include bloggers, YouTubers, social media influencers who produce engaging tutorials, investment strategies, and market analysis that resonate with a financially curious audience."

Matt’s reaction was "nailed it," and the point he made right after is the one that matters: that is the several hours of bullpen brainstorming, done in about as long as it takes to read this sentence.

Network, affiliate signup link, commission, cookie duration, any bonuses, and your email signature. We put in 20 percent and a 90 day cookie. These are not admin busywork. They become merge fields in the outreach, so the affiliate sees real terms in the first email instead of "let’s hop on a call."

This is where you get specific. Alfie picked ten keywords and suggested more. It assigned a category from roughly 700 IAB categories, which are the industry standard rather than something invented in-house. Then you toggle the affiliate types you actually want.

Matt’s take on that screen is worth quoting, because it is the question every brand asks an agency on day one:

"The question upfront of any program that comes and tries to bring on an affiliate manager or an agency to do a recruitment always wants to know, well, what type of affiliates can you recruit for me? Maybe they say I don’t want coupon sites or they say I do want coupon sites. We’ve taken thirty years of knowledge of knowing what type of affiliates live out there, who drives actual traffic, and we’ve trained Alfie to go find the type of affiliate that you are looking to find."

Layer on geo, age range, audience size, and gender if you want them. We set United States and left the rest open.

Alfie searched two places at once: a data lake of roughly 80,000 pre-verified affiliates, and the live web.

That second half matters more than it sounds. Beth was honest about why the data lake exists in the shape it does. The team had decades of unstructured affiliate data sitting around, and turning it into something usable meant finding new contacts for people who had moved and dropping sites that were not even live anymore. In her words, "affiliates are pretty finicky, they come and go, and they move around a lot."

Matt connected that to the thing every affiliate manager has quietly noticed about network marketplaces:

"The network’s data and their marketplaces, the data is only as good as what’s been inputted by the affiliate themselves. If that person that owns that account leaves and that data changes, you don’t really get that appended data or verified data."

He also pointed out that there used to be vendors walking around Affiliate Summit selling spreadsheets of affiliates. Data brokers, basically. Long gone are the days of buying a static list at a conference, because a static list starts rotting the day it is printed.

Click any affiliate and a drawer opens with a summary, the reason Alfie picked them, the contact or contacts it found, a fit score, promotional types, geo reach, traffic and authority numbers, and soft signals like audience age and income. On the Robinhood run, one match came back 96 percent US traffic with a fit score of 90 out of 100.

Beth made the point that this is the SimilarWeb tab and the SEMrush tab and the contact-finder tab, all collapsed into one drawer. One of our users, Jamar, said in the live chat that he is currently stitching three tools together to do this. Matt said he keeps three screens open at all times for the same reason.

Approve the good ones. Reject the bad ones with a reason, which trains the system for your brand. Rejected partners go to a pile you can pull back from, so you never lose anyone permanently.

Approved affiliates drop into an outreach sequence, which works like an autoresponder. Three emails: the first on approval, a follow-up three days later, a third six days after that. The sequence stops the moment someone replies.

Fill in any merge tags Alfie flags as empty, hit queue and send, and it goes. By default it sends from alfiemail.com with a tagged address so replies route back to the right campaign and affiliate. Set up your DNS and it sends from your own domain instead.

Matt told a story on the episode about Kristen Evans, who runs an agency. She set a campaign in motion and then went and got a run in on her treadmill while Alfie was out finding the affiliates. That is the actual unlock. Not that the work got faster, but that it stopped requiring you to sit there.

Replies land in an inbox view. You can write your own or have Alfie draft one and then tell it to make the reply more casual. Move interested partners to Engage, then Deploy when you are sending links and creative, then Convert.

Matt’s framing: this is not a discovery engine, it is the life cycle of an affiliate from first contact to active partner.

One honest note: there is no export today. Your partner list lives inside Alfie. Beth said the team keeps it contained so the workflow does not lose visibility, and they are looking at integrations. I said on the episode that a CSV import will eventually be on the roadmap. It is not there today, and I would rather tell you that now than have you find out later.

A fit score is a single number, out of 100, that tells you how well a discovered affiliate matches the campaign you set up.

Beth broke down what goes into it: confidence in the data, whether the keywords match, whether the category matches, and whether the site meets the geo and demographic criteria you set. Does the system have complete information on this partner, and does their actual content line up with what you asked for? (She also noted that Dan, our CTO, sent her the full algorithm right before the call and it was over her head, which is the most honest thing anyone said all hour.)

Here is my take, and I do not hand this out lightly. I have been working with tools in this space for the last ten years, going all the way back to FiveIQ and MediaRails. This is the most sophisticated search functionality I have ever seen, and the fit score is not something you can easily replicate. You could go out today and ask Claude or ChatGPT for affiliates and they will give you some answers. But the fit score is incredible. It is spot on when it is hunting down the right type of affiliate for your brand, the kind that is actually going to be effective.

There is a second signal next to it: the verified badge. That means the system looked at the site and found the markers of a real affiliate, things like live affiliate links and affiliate disclaimers. Unverified is not automatically bad. Beth pointed out that some clients specifically want brand new content sites that are not affiliates yet. Just know the conversation is different, because you have to teach them how affiliate links and programs work before you can talk terms.

Real numbers from the episode, straight from Beth Lazar:

Matt Frary added that inboxing rates are running in the high nineties.

If 3.8 percent sounds low to you, sit with it for a second. This is cold outreach to people who have never heard of the brand. Compare it to the reply rate on the emails you are currently sending to info@ addresses, which is functionally zero, and the math looks different.

The proof point Matt gave is my favorite from the whole hour. Their services team, Chief of Chaos, has been running Alfie for client programs, and they got the New York Times to respond to an up and coming brand through one of these sequences. Not a household name brand. A new one.

Yes, and the honest answer is that it surprised Matt too.

He has run a B2B agency, and he expected the data to skew consumer. Instead the reporting showed the affiliate data skewing heavy toward B2B. His best argument is the most self-referential one available: Alfie is a B2B SaaS product, and Alfie is out there recruiting affiliates for itself, at a 7.9 percent response rate.

On my side, I have built a lot of demos on this thing, and it finds partners for any niche no matter how small.

Two limits worth knowing before you sign up:

Here is the short version. Finding affiliates has been a manual grind for twenty years because nobody could automate the messy middle: the searching, the qualifying, the contact hunting, the first draft of the email. That part is now automatable. The relationship part, the negotiation, the strategy, the reason a partner picks you over the 10,000 other requests they got today, that part is still yours and always will be.

This is an enhancer, not a replacement, and that is where this industry is leaning. You have got to get on board.

If you are still building your partner list out of a spreadsheet and a Google tab, pick one brand this week and run it through a tool. Any tool. Time how long it takes you to get eight qualified affiliates with real contacts and a sent email. Then compare it to your last recruitment push.

Try Alfie free for 14 days. No credit card required. Start at alfie.io

What is the slowest part of recruitment for you right now? Finding them, reaching them, or getting them live? Tell me, because I am genuinely curious which one is the real bottleneck for most of you.

How do I find affiliates for my brand?

Start with the keywords a buyer would search before purchasing your product, find the sites and creators ranking for those terms, verify they run affiliate links and get real traffic, locate a working contact, and send a personalized pitch with your commission and cookie terms in it. Affiliate recruitment software collapses those steps into one workflow that starts from your URL.

What is the best way to find affiliates without buying a list?

Search live rather than buying static data. Contact information on purchased lists and network marketplaces goes stale fast, because affiliates move around and marketplace profiles are only as accurate as what the affiliate typed in themselves. A tool that verifies contacts and searches the live web will beat a spreadsheet you bought at a conference.

What response rate is normal for cold affiliate outreach?

On Alfie, cold outreach across new users over a recent 38 day window ran at 3.8 percent, the Alfie program’s own recruitment ran at 7.9 percent, and one client hit 25 percent. Inboxing rates were in the high nineties. Rates vary by vertical, offer strength, and how specific your pitch is.

Does AI affiliate recruitment replace affiliate managers?

No. It removes the search, qualification, contact hunting, and first-draft writing. Negotiation, relationship building, strategy, and optimization stay with the human. The point is to hand the manual work to software so the manager can do the work that actually needs a person.

Can I send affiliate outreach from my own email domain?

Yes. By default Alfie sends from alfiemail.com with a tagged address so replies route back to the right campaign. Set up your DNS and outreach sends from your own domain instead.

Does this work for B2B and SaaS affiliate programs?

Yes. The data skews heavier toward B2B than the team expected. Alfie is itself a B2B SaaS product recruiting its own affiliates through the same system, at a 7.9 percent response rate.

Can I export my affiliate list?

Not today. The partner list stays inside Alfie so the workflow keeps full visibility on what happens after outreach. Integrations are being evaluated, and a CSV import is expected on the roadmap, but neither is available right now.

Does Alfie have an API?

Not yet. An MCP is in the works. Talk to sales if you want an early preview.

Recorded live on the School of Chaos podcast with Matt Frary, President and COO at XPFlow, and Beth Lazar, Chief of Product at XPFlow.