GTM struggles of 2025: Experimenting your way to GTM fit

Startups need to find GTM fit quickly. 

‍

We're in a race against time. Funding is running out. Competition is lurking. 

We want to run experiments that are statistically significant, because I don’t have a quarter to lose. I want to draw confident conclusions based on data.

‍

Let’s talk about how we can do this more quickly and cheaply than ever.

‍

Disclaimer: these automated, experimental campaigns are not going to be your best converting campaigns. 

‍

They aren’t going to have fancy touch points in there like direct mail or phone calls. 

‍

We know that we can improve our results if we add touch points like this. But right now, we're looking to benchmark how each segment performs to work out which one we should double down on.

1. Building and filtering your company lists

The first step is to build a list of companies and contacts. These are your segments.

‍

You might do this in GoodFit or another data provider. You can do it in your CRM if you've already got lots of accounts in there that are rich with up-to-date data. 

‍

You then start to draw and filter for lists of companies that might meet each one of your segments or your hypothesis. And importantly, you want to try and ensure those companies are a true match for your hypothesis.

‍

We’re not in the territory of ‘software companies that have raised over 10 million’ or ‘founded after 2010’. These are not specific enough groups. If you find yourself describing them only in attributes, you're probably not doing it right. 

‍

There needs to be a commonality of pain or need. 

‍

Now we need to sense check the size of our sample. 

• Is it big enough to warrant a programmatic approach in the first place? If it’s 50 companies, don't bother. It doesn't make any sense to set that up programmatically. 

• Even more importantly, our end goal is GTM fit - so if you can only find 50 companies that fit neatly into one of the segments you’ve hypothesised around - that isn't going to be big enough to repeatedly go and sell into anyway. 

Now if you set up your companies and filters and you get back a few hundred of them - that’s a sign you’re on the right track.

‍

I’ve been asked what the minimum size is, and I think it's probably in the low to mid hundreds. 

‍

What we're looking for is what I call evergreen segments - a set of companies you’re going to move into as they emerge. It doesn’t matter if you’re at the lower end of those hundreds, because new companies are going to emerge over time. It’s pointless doing all this segmenting then not having anywhere to go next. 

‍

You can also recycle companies and contacts within a segment. I don't want to work a company more than once every six months or so. But if I’ve got 400 companies, I can do 400 companies over two quarters and then recycle them in quarters 3 and 4. 

‍

Example:

GoodFit is a data provider. In theory, we can sell to any B2B sales org out there, but I have a hypothesis that we're uniquely positioned to win the B2B sales orgs out there with complex territory planning requirements. 

‍

Normally, they need lots of accounts of different sizes and different regions, and this is great for us because they aren't doing self prospecting. 

‍

Even better for me is if those complex territory requirements are owned by someone who understands the pain - i.e. RevOps. This is my hypothesis for the type of company that GoodFit is uniquely positioned to serve.

‍

If I want to draw an audience around that segment of customers, I filter for:

1. Sales team count between 3 and 150 (minimum 3 is really important to me)
2. 1+ people in RevOps 
3. “Territory planning” keywords (if someone's running territories, they tend to mention it in sales manager, reps, and RevOps job descriptions). 

I can now go out there and deliver them a message on how GoodFit is uniquely positioned to help companies just like them. 

Maybe my hypothesis will be proved right, maybe it will be proved wrong. 

‍

Another example:

‍

My previous company was Paddle, which does payments for software companies that sell via a checkout, credit card, or online payment.

‍

We had a hypothesis that proved to be true - Paddle was uniquely positioned to help software companies who had a need or a desire to sell their product globally. We were really strong at international tax, currencies, languages, and payment methods. 

‍

So how do I draw a segment of folks that have that need or pain? I would filter for software companies with:

1. International traffic of greater than or equal to 20% 
2. Fewer than two currencies supported on site
3. Fewer than two languages supported on site

So these companies are getting traffic from all over the place -  South America, Europe, India - but they’re only supporting GBP, USD, and a single language! They’re not optimized for the traffic they’re already getting. 

‍

I found over 20,000 companies meeting this criteria. That's actually pretty huge - I'd probably look to decrease it, perhaps by upping the international traffic threshold.

‍

Remember there will also be new companies emerging all the time that pass my thresholds. So it's certainly a segment I can sell into on an ongoing basis. 

‍

And because I can sell into it, if I see success, I can invest so much more. I can build landing pages around this pain. I can run webinars on this pain. I can invest a hell of a lot more in targeting this group with more advanced touch points and mechanisms. 

‍

I have talked in great detail about setting up programmatic outreach, so would divert you to this webinar if you haven’t seen it already.

‍

2. Messaging that reflects the need

Next, set up a sequence or a message that reflects your ability to solve the pain and the need of those in the segment. 

‍

This is probably quite obvious, but if I've just built a segment around people that have a load of international traffic and no currencies or languages for it, talk about that in the message that you send them!

‍

How easy it is to define this message is also a good litmus test for whether your segments are tight enough.

‍

All you need to do is draw circles, ideally cross-functionally/with some other people in the org, around the companies and contacts you want to run tests on over the next quarter. 

‍

3. What results are we expecting?

Nothing huge. 

‍

A basic, completely automated email and LinkedIn sequence is not going to drive a 15 to 20% response rate. 

‍

But, what we're actually looking for is the green shoots of the segments that are worth investing in.

‍

Say we run 3 experiments in a month that get a 0.5% response rate, but another 2 of our experiments get 3%.

That's certainly within the realms of possibility. It's those successful experiments we want to double down on. 

‍

And now you have some decisions to make. Maybe you'll continue running programmatic outreach to a broader group to see if the performance holds. Maybe you'll do that with some tweaks to the sequence or the message. 

‍

Maybe you'll layer on cold calls, because you know they’re more likely to take an initial call. Maybe you try running direct mail at them or invite them to a webinar or an event.  

‍

All we've done is narrow our focus on a group of customers that we've got proof points and data to suggest that we're probably a good fit for. 

‍

This method helps you quickly, easily establish where to invest - as opposed to the past where you'd be hiring people, taking punts, and throwing out messaging that isn’t well designed.

‍

The next step is to continue to iterate in the segments you're having success in. Continue to experiment into new customer groups. Your level of GTM fit is going to vary across segments.

‍

To use GoodFit as an example, one hypothesis might be: we help people with territory management. Another might be: we help people with programmatic outreach. Another could be: we help people in FinTech selling to FinTech, because it's a really hard data set to get your hands on.

‍

I'm going to deliver a different message to each of those groups - that's to be expected. But I also want to run a similar automated sequence, in style, in tone, in length, in touchpoints, in time - keeping that stuff as constant as I possibly can. 

‍

Because we're trying to see which customer type we're most likely to succeed with. 

‍

We're not yet at the stage where we're iterating or optimizing for the sequences themselves, or trying to maximize our initial call rate. That's a very different type of experimentation that we want to do only once we’ve established our groups. 

‍

4. What not to do

‍

1. Contact so few people, you can't draw any conclusions whatsoever. 

‍

I’ve seen founders trying to establish GTM fit in a more manual way, and when asked how many inboxes they’d hit or people they’d got in front of, they’d say like eight. You’re never going to draw any meaningful conclusions from that level of output.

‍

2. Target all sorts of different-looking companies with different needs, or put out all sorts of different messaging.  

‍

If I’m sending every single company in my segment a slightly different message with a slightly different CTA, a slightly different subject line… I'm introducing way too many variables. Try and keep it systematized. 

‍

We’re trying to get to a point where we can confidently say we've reached out to groups A and B, and we're seeing better results for group A. 

‍

Then: here's how we're going to iterate within that group, or continue to invest in winning them whilst running a new experiment into Group C - and do so with statistical significance and confidence. 

‍

TLDR: Take your segment, enrich contacts, bring them into a tool (Lemlist is an option) that allows for email and LinkedIn automation, and set up a sequence. Then you can start enrolling folks on an automated basis in their 1000s.

‍

The best part? 

‍

You can do all of this in GoodFit. Speak to our team today!

‍