Startups are hard. I know because I’ve spent over a decade working at them.

And early-stage startup marketing is one of the hardest parts. Not only are you trying to get customers but you’re also trying to figure out whether there’s even a market for the product you’re selling.

I’ve experienced highs, lows and a lot of hard-earned lessons. This is a list of a few of them, as a reminder to myself to learn from them.

1. Once a channel works, go all in.

In 2015, I worked at a company called Digital Mums. We sold vocational social media marketing training courses to mums in the UK. We weren’t quite bootstrapped since we’d taken a £50k convertible loan, but we weren’t far off.

It was scrappy, it was fast-paced and we had minimal marketing budget.

I started experimenting with Facebook Ads. My first experiment generated 50p a lead. The courses we were selling we we worth around £500 and increasing in price as we got more customers.

My first reaction should have been “holy shit, I need to pour everything we’ve got into these.” Unfortunately, my actual first reaction was along the lines of “cool, I can slowly increase this as we get more budget and have this be a sustinable channel for us.” That was a mistake.

A year later, that price had 4x’d and then a year after that it had 2-3xd again. It still became a sustainable channel for us, but I could have 3x’d the company’s growth at that time if I’d gone all in on it. If something works you have to scale it as fast as possible. Growth cures everything and marketing channels don’t last forever.

2. Cold outreach works.

The model at Digital Mums worked like this: every month we’d run a training course. We’d sell x number of spaces to mums. These mums would then run a real social media campaign for a real business. That meant every month we had to find each mum a corresponding business. Sell 10 course spaces, we then need to find 10 businesses. And since sometimes we’d sell a course space a week or two before it started (we didn’t have the luxury of turning revenue down) we’d have a mad rush to find a business in time for the start date.

I’ve done two-sided marketplaces three times in my career. Every time, it’s challenging. What works for one market does not work for the other, so you need two completely different go-to-market strategies, but you have the same amount of internal resources (usually just me) as if you were targeting just one.

I really didn’t want to cold calling and cold outreach. Manual as hell, semi-humiliating at times, not particularly interesting our scalable.

But, you know what, it worked.

I cold-called co-working spaces. I dug up lists of charities and small businesses and painstakingly emailed them one-by-one. Not the most fun I’ve ever had, but effective. We had some close calls and we didn’t hit targets every month, but cold outreach proved to be one of the main lifelines for this side of the business.

3. But it depends on the right offer from the right person.

I’ve done some outreach that worked and some outreach that didn’t work. Whether it works comes down to two main things from what I’ve seem.

a. The sender

Are they positioned in a way to make them interesting to the recipient? Do they have some sort of authority or social proof that ‘vouches’ for them?

At (Acapela)[acape.la], we pivoted the business model and targeted Product Managers. We did cold outreach via LinkedIn to try and get PMs to beta test the newly pivoted product. Our Founder Roland had previously sold a company to Reddit, so we did the outreach through him and threw in Founder of Dubsmash (aquired by Reddit) into his LinkedIn title.

Over 50% of leads connected with him. Response rate was like 33%. Call booking rate something like 10-15%.

And even better, because so many connected with him on LinkedIn, any time he posted on LinkedIn organically, PMs were seeing his posts for free.

b. Whether it’s a good offer.

At the beginning of Digital Mums, we were selling businesses a four month social media campaign for £99-199 (because the social media marketer was in training). The payment was really more of a way to get buy-in from than business than any sort of real revenue at the beginning. It worked out to something like £1 per hour of work done. That’s an amazing offer. The cold outreach worked as a result.

Tricks and clever subject lines don’t matter. Is the person messaging of interest and is the offer good (and relevant)? That’s 90% of the work.

1. It’s sometimes easier (and better) to go mass-market rather than trying to niche down.

The general advice is identify your ICP, then hone down and only target them. Obviously, you should target relevant potential customers. But you shouldn’t be afraid to go mass market.

One of our first big wins at Digital Mums was getting an article published in the Mail Online. In the UK, that’s about a mass-market as you can get.

We drove something like 2,000 leads from it, a lot of those who were not our target audience. But among those 2,000 a percentage (say 10 per cent) were the right ICP.

That was like 200 targeted leads, which would have taken weeks at the time.

And those other 1,800 leads weren’t just waste. They were people who now knew about Digital Mums and could tell other people about it.

There was a very clear before and after that article.

More eyeballs is nearly always a good thing. Sure, hone your messaging and your positioning to target your ICP’s painpoints. But going mass market can still mean you target the right people AND you get better amplification.