200 startups

Over the last twelve months I've roasted the landing pages of 200 startups. Landing pages from indie makers, VC-funded scale-ups and profit-generating enterprise organisations, across multiple industries and audiences. With an average of 25 mins a roast, that's over 3.5 full days of roasting pages to increase conversion.

What is a roast?

Each roast was a personalised 20-minute video review of a landing page, identifying fixes to convert more visitors into customers.

Based on best practice, testing insights, and my work as a conversion rate optimisation expert.

But what did I learn about high-converting landing pages? And about running a productised business? And what can you learn from me? It turns out quite a lot...

Why I started Roast My Landing Page

Roast My Landing Page started as a pandemic side-project to allow me to support early-stage startups while generating a second-stream of income outside of my freelance marketing work. It ended up generating me over £70,000 of work, over 2,000 email subscribers, and an ever-growing list of insights into the pains and challenges of early stage businesses and the amazing people that run them.

It's all about the founders

And it has really been about the founders! Every single person comes with their own story, passion, and ideas. They were eager to learn how to do justice to their product, service or email list, and showcase it to the world in the form a high-converting landing page.

This post is for every founder who thanked me, challenged me and (in rare cases) fought me. It's been a blast. So, here's what I've learnt...

The 9 most common (and easily fixed) things that founders miss...

One goal

50% of founders had one clear goal for their landing page - sign up, download a lead magnet, or book a demo. The other 50% had multiple, often equally prioritised call-to-actions. This often leads to analysis paralysis and confusion for the visitor.

Fix: Focus your landing page on one conversion goal.

A focus on USPs

After booking a roast, customers of Roast My Landing Page are asked to fill in a quick questionnaire. One of the questions is "what makes your business unique?" Nearly every founder was able to capture their product or business USPs gracefully in the form, but only about 1 in 5 had this language on their landing page.

One of the key take aways from your landing page needs to be what makes you unique. If a buyer is in the consideration stage (comparing solutions) your USP is what is going to make them recall you, and decide if you are the best fit for their needs.

Fix: Contrast your product with competitors and the current way of doing things.

Clear, relevant social proof

Testimonials, ratings, awards or quantifications to evidence that other people use and love your product. Only about 40% of landing pages had it above the fold (what people see when they first land on your page). And of those, only about 50% used testimonial language that was consistent with the pain and benefits listed elsewhere on the landing page.

Fix it: Move your social proof up the page, make sure it's concise, compounds your copy, and from a buyer relevant to the visitor.

Simple language

About 1 in 8 landing pages I read, I couldn't comprehend on first read-through. These landing pages required several attempts to simply 'get it'. I had a monetary incentive to try and work out what was going on - your visitors do not.

Fix it: Avoid technical terms and acronyms, and write in plain language. Remember you're talking to a human, even if they are a B2B buyer. Ask yourself if a 12 year old could understand your landing page.

Real pain

PAS (pain - agitate - solve) is a common copywriting technique used to increase conversion. Most landing pages touched on the pain they were addressing, but only 1 in 15 agitated or amplified the pain with emotional language and vivid imagery. The ones that did this well created much more powerful landing pages that moved me to explore the solution.

Fix it: Agitate your visitor by painting a vivid picture of the pain using emotional language, stories and visuals.

Clear benefits and use cases

Lots of pages talked in-detail about product functions but ignored benefits and use cases. Research consistently shows higher conversion with benefits-based language. It was left to the visitor to think how the product would benefit them, or specifically solve their problem.

Fix it: Visitors shouldn't have to work out how and why the product will benefit them. Tell and show them with clear benefits language, and example use cases.

Signposted CTAs

So many landing pages had a call-to-action to do something, without clear signposting around the action. Yes, they were they about to sign up but... how long would it take? What were they agreeing to? How much would it cost? What technical setup was involved?

Fix it: add context to your CTA so the visitor knows what to expect.

Asking instead of over-thinking

3 in 5 founders mentioned a level of frustration over not knowing what was happening during the pre-roast survey. They weren't sure why their landing page wasn't converting visitors into customers. However, very few had actually asked their visitors.

Fix it: sign up to GetSiteControl and add an exit intent survey on your landing page. Asking why they're leaving. Address those reasons on your page.

Knowing their stats

Only 2/5 founders could tell me current conversion on their landing page. In most cases these people didn't even know if they had a conversion page issue but had booked a roast anyway. Most founders had Google Analytics installed but had limited goals tracking setup, or hadn't reviewed reports in months.

Fix it: setup analytics and events tracking. Uncover a conversion baseline for future experimentation.


7 (slightly more) advanced ideas for better performing landing pages

About 1 in 4 bookings were to roast landing pages from scale-ups where there was clearly a high level of expertise and skills used to build the landing page. These businesses still had conversion challenges, which fell mainly into the following 7 areas.

Your landing page isn't niche enough

I saw this over and over, and was asked about it as often. "Should I target a bigger audience or a smaller one? Surely the bigger the market the better?" My advice was always them same. Build a landing page for the most niche audience possible until you've secured your first customers. Build out more landing pages later.

Landing pages targeting multiple personas and use cases almost universally convert at a lower rate. Go more niche, expand later

Your ask is too big or too soon

Nobody is signing up without understanding their own problem, your solution, and your pricing. And very few people are paying an enterprise SaaS without speaking to an account manager first. Be mindful about what you ask - your CTA. Too soon on the page or in the buying cycle, it's not going to convert.

Write a list of what a visitor needs to know before they will sign up. Make sure this critical info is before your first CTA.

You're telling people things you can show them

So many pages used large blocks of text to explain something that could be more clearly and powerfully demonstrated in a simple product shot, table, visual, example, demo, illustration or abstraction.

Show, don't tell.

You're not addressing doubts

On visiting your landing page, your potential buyer will be forming questions in their head. As these unanswered questions grow, the chances of a conversion decrease. Find out what they are, and address them throughout the copy, or in a FAQs module.

Understand and address doubt through user testing, exit intent, or surveys.

You need to use better images

Around 65% of the landing pages I reviewed used images from popular landing page collections: photos, icons and illustration. Although they were workable and even at times relevant, they rarely told the same story as the copy visually.

Work harder to find or create meaningful images for your landing page.

You don't know your stats

Even people who were extremely smart and knew their landing page performance, had not measured their full funnel from marketing channel source, to account sign up. This meant they would often have a fantastic landing page but sign up pages funnels that weren't measured.

Setup funnels in Google Analytics or other tools so you can review drop off throughout the funnel, not just on your landing page.

You need to implement regular conversion testing

And finally. You may already have a high-converting landing page, but without regular testing you're not learning. No customer had a regular schedule of tests in place. At best, they had last checked performance after relaunching their site several months previously.

A cycle of constant experimentation increases your learnings about customers and drives more revenue. Keep shipping tests!


Things I learnt about building a productised business

💸 Making money

  • I generated about £20,000 in roast revenue, and another £50,000 of freelance marketing work from clients who discovered me through being roasted.
  • I spent about £7,000 in paid performance advertising and about £3000 in tools and other business expenses.
  • I've recently added follow-up services - copy rewrites, landing page rebuilds, conversion analytics implementation. Around 1 in 4 of my roast clients go on to book a second service.
  • About 1 in 6 clients offer me freelance marketing work.

😱 Founders and conversion

  • The most common reason for booking was to get a second opinion based on a sense their landing page wasn't working, or their positioning was wrong.
  • Most founders did not know their current landing page conversion, even if they had the analytics in place.
  • Nearly every single founder used Google Analytics to measure site analytics, most had set up Goals, but many didn't actually review their performance.
  • 95% of roasts were booked by male founders.
  • The most common landing page page types were SaaS, ecommerce and B2C apps and tools.

👀 Competition

  • Around 5 makers built near-identical products that were clearly inspired by Roast My Landing Page.
  • Many lifted copy, concepts, CTAs, pricing and marketing ideas directly from my the Roast My Landing Page website.
  • Others were inspired but added their own spin.
  • Some reached out and asked for my support, feedback and advice, which I tried to offer.
  • Many of these projects seemed to close within weeks.
  • I made a point to not focus on what they were doing and instead put the energy into growing my business. But there was some moaning to friends.

📣 Collecting feedback

  • Before I roasted a landing page, I sent the customer a form with 7 questions to complete - to help them reflect and focus, and give me context.
  • Post-roast I sent all clients an anonymous Typeform survey, asking them one idea to improve the roast, which I used to refine my offering.
  • Approximately 1 in 4 completed the survey, 1 in 4 emailed feedback and 1 in 2 I didn't hear from again.
  • The survey included a rating out of 5 stars. Of the people that replied, I received 46 5-star ratings and 4 4-star ratings.
  • The survey also asked for other parts of their business the founder would like roasted - this lead to me creating other services.
  • Around 25% of customer replied to my suggestions personally.
  • About 50% of founders implemented the fixes detailed in the roast. Most either did so immediately, or several weeks (or even months) later.
  • More recently, 1 in 4 customers have asked me to complete the changes for them.
  • The more I built in public, the more help I got from strangers. People would regularly send my ideas, fixes, even their own mini-roasts, multiple times a week.

✅ How I acquired customers

  • The Roast My Landing Page landing page had various conversion issues.
  • I iterated on the landing page about 20 times, tested 3 payment solutions, 4 price points and 5 buying flows. I am still testing it now.
  • My best performing marketing channels were word of mouth, participating in the Indie Hackers and Productize communities, Twitter Paid, Facebook Paid and then Google Ads, partnerships.
  • The best ROAS on paid ad spend was about 3.
  • The worst performing marketing channels were Quora, Reddit, emails through Paved.com. Multiple tests through these platforms failed to deliver meaningful conversion.
  • The best way to generate word of mouth was to overdeliver. I promise to return each roast within 48 hours, but many customers received theirs within an hour of booking. This created a mind-blown moment for the founder and drove significant word of mouth.

🛠 Tools I used

  • Tools like Loom are amazing for recording video-in-video. But when they fail mid-roast recording it's horrible. When it happens 3 times in a row, you consider leaving your job, hiring a small dependable car, packing up your possessions and moving to a remote farm far away from everything and everyone you know.
  • I used SPP to manage bookings.
  • Typeform for customer feedback.
  • Google PageSpeed Insights for page load times.
  • Autopilot for my mailing list.
  • Ghost for this blog.
  • Analytics tools I used included Hotjar for recordings and heatmaps, Heap for funnels, and Google Analytics for high-level reporting and conversion indicators.
  • GetSiteControl helps me collect leads to nurture through my weekly email.
  • Hotjar helped me collect exit intent insights.
  • I used Stripe and Paypal for payments.

🤑 Pricing my product

  • A roast started at £39, went up to £299, and is now £199.
  • I tested a few different price points to identify what would drive the most profit.
  • Filming a roast requires mental agility and is extremely tiring, so my limit was about 4 a day.
  • I moved to value not time-based pricing as I built evidence of the roasts increasing conversion.
  • I am still struggling to attract the right clients at the right price point without ending support for bootstrapped makers.

🪜 Processes I implemented

  • Routine is good but I struggle with it: weekly cycles of reporting and personal retrospectives kept me on track.
  • Yes, I ran retrospectives with myself focusing on what I could improve from the previous week.
  • Accountability helped me focus, especially building in public on my Twitter and detailing my Milestones on Indie Hackers.
  • I also have a small Slack community of peers to share ideas with, and stay accountable.
  • The roasts themselves are relatively freeform. However, over time a loose structure evolved that ensured I prioritised the biggest opportunities.

What's next?

Not only is Roast My Landing Page profitable, but running the business has been rewarding in a ways I hadn't anticipated.

I'm not (yet) sat on a beach sipping margaritas passively earning mega bucks, but I am enjoying sharing actionable feedback for a fair price, seeing my clients' businesses grow, building my network, and continuing to improve my understanding of founders, their businesses and conversion optimisation.

Mid-April 2021 I went full time on roastmylandingpage.com and share regular updates as part of #openstartups #buildinpublic on my Twitter.

I also video roast one SaaS landing page a week in my free newsletter, which I'd love you to join.

Thanks to everyone in the Productize and Indie Hackers communities, and on Twitter for their support over the last few months ♥️

Want to turn more visitors into customers? Book a roast and get £10 off with the code 200roasts.