When was the last time you actually used what your customers told you to write an ad?

I mean really used it - not just nodded along in a call and then went back to your desk to write what you think sounds good.

This week I want to talk about something we're seeing work brilliantly across the businesses we work with: using real customer language and pain points to power your ads. And specifically, how to test this approach in new channels where it really matters.

The catalyst for this? We've been helping a B2B SaaS client test Reddit as a channel to reach technical decision-makers. And it's been a masterclass in why customer insights are your best creative brief.

Why customer insights make better ads

Here's the thing about most B2B ads: they sound like ads. They use the language of marketing departments, not the language of real people having real problems.

But when you base your ads on actual customer conversations, something magical happens. Your ads sound like they're part of the conversation your audience is already having, not an interruption to it.

Take this example from our customer insights work: when we ask "What was going on in your life that brought you to [product] today?", we're not looking for polite feedback. We're looking for the messy, real reasons people actually buy things.

One of our clients discovered through these conversations that their customers weren't searching for "cloud cost management platform" (their product category). They were searching for solutions to "how to stop AWS bills spiralling" and "getting engineering teams to care about costs". Completely different language. Completely different ad strategy.

The Reddit case study

Reddit is interesting because it's brutally honest. If your ad doesn't sound authentic, you'll know about it immediately (and not in a nice way).

We're working with a B2B SaaS company that provides developer tools. They wanted to reach CTOs, engineering leads, and technical architects - people who shape vendor decisions. These aren't people who respond to generic "transform your business" messaging.

Here's what we've done differently by starting with customer insights:

Started with real pain points

Through customer conversations, we identified that these technical buyers struggle with three specific things:

• The build vs buy decision (not just "should we use this tool")
• Proving ROI whilst reducing engineering time
• Maintaining security and compliance whilst moving fast

These aren't marketing assumptions. These are verbatim themes from dozens of customer calls.

Used their actual language

Rather than talking about "innovative solutions" or "next-generation platform", our Reddit ads use the language we heard in conversations. Things like:

• "We've seen engineering teams reclaim 15 hours per week by switching from building in-house"
• "Deploy faster. Enterprise-grade security included. No DevOps team required"

Notice the specificity? That comes from real conversations, not a thesaurus.

Matched the channel

Reddit's magic is in conversation-style ads that appear directly in comment threads. So we created ads that sound like a helpful colleague joining a technical discussion, not a sales pitch. Because that's exactly what our customer interviews revealed these buyers respond to - peer recommendations, not vendor hype.

One of our test ads literally starts with a question someone asked in a customer interview: "How are you managing integrations across your stack without a dedicated platform team?" Then answers it like a Reddit user would, before introducing the product. It feels native because it is native - to the actual conversations happening in that space.

How to do this yourself

Here's the practical framework we use to turn customer insights into ad creative:

Collect the right insights

You need to ask questions that uncover real motivations, not polished answers. Some of our favourites:

• What else have they tried and why didn't it work?
• Where do they go looking for answers?
• What stops them from trying something new?
• What was the specific moment they decided they needed to solve this?

Timing matters too. Ask immediately after sign-up when it's fresh, not six months later when they've forgotten.

Categorise and prioritise

Don't just collect feedback - analyse it properly. We use AI to help with this (especially useful when you've got 1000+ survey responses).

Here's a prompt we use constantly:

"You are a conversion copywriter expert in persuasion. Review these customer responses and identify the top motivators. Create categories, rank them by frequency, and provide exact quote examples for each."

This helps you spot patterns you'd miss manually. Maybe 60% of your customers mention "reducing engineering time" but you've been focusing your ads on "cost savings". That's your signal.

Map to the customer journey

Different insights work for different funnel stages. For Reddit, we're using:

• Awareness: "Real language" about pain points technical audiences are actively discussing
• Consideration: Proof points from similar users ("Engineering teams at companies like yours are saving X hours")
• Decision: Specific outcomes and "build vs buy" framings

Test ruthlessly

Don't just create one ad and call it done. Create variants testing:

• Different pain points (technical debt vs time-to-market vs team capacity)
• Different proof points (customer quotes vs data vs outcomes)
• Different formats (conversation-style vs direct vs educational)

For our client's Reddit test, we're running two creative angles:

• A: ROI-focused - "Deploy integrations faster. Save 15 engineering hours per week. No in-house build needed"
• B: Thought-leadership - "Why smart engineering teams are choosing to buy, not build"

Both based on different insight themes from customer conversations.

Why new channels need this even more

When you're testing a new channel like Reddit, you can't rely on what worked elsewhere. Different platforms have different cultures, different expectations, different BS detectors.

That's why customer insights are even more critical. You're not guessing what might work on Reddit - you're testing messages you already know resonate because they came from your customers' mouths.

Compare this to Google Ads, where you're competing with anyone searching a keyword (students, researchers, competitors). On Reddit, you're showing up in discussions where practitioners are already engaged. But only if your message sounds like it belongs there.

Think about ads that work on Reddit - they feel like they're part of the conversation, not an interruption:

• Using customer quotes that sound like Reddit comments
• Creating problem-solution framings that mimic how users actually help each other
• Framing content as educational guides rather than sales pitches

These work because they're built on understanding how these audiences actually talk and think.

What this looks like in practice

Here are some before and after examples from our work:

Before (marketing language): "Enterprise-grade integration platform for modern development teams"

After (customer language): "Tired of maintaining integrations? We handle the OAuth, webhooks, and API versioning so you don't have to"

Before (generic pain point): "Struggling with inefficient workflows?"

After (specific pain point from research): "Spending 2 days a week managing customer data across 7 different tools? There's a better way"

The difference? The second versions use language and specificity that came directly from customer conversations.

The mistakes to avoid

Using customer insights sounds simple, but here's where people get it wrong:

Mistake 1: Cherry-picking the insights that fit your existing strategy If customers keep mentioning price but you want to position on innovation, you can't just ignore it. Let the insights change your strategy, not confirm it.

Mistake 2: Using "sanitised" versions of customer language If a customer says "I was drowning in integration hell", don't change it to "seeking operational efficiency". Use their words.

Mistake 3: Only using positive feedback The stuff that didn't work for them (other solutions they tried) is goldmine material. It tells you exactly what to differentiate against.

Mistake 4: Treating all insights equally Not every comment deserves an ad. Focus on the patterns - the things you hear repeatedly across different conversations.

The results we're seeing

It's early days for our client's Reddit test (we're only two weeks in with a modest budget), but the signals are promising. We're seeing CTRs above 0.5% and strong dwell times (60+ seconds), which suggests the messaging is resonating.

More importantly, we're learning exactly which pain points drive engagement in this audience. That's insight we can use across all channels - LinkedIn, Google, even sales outreach.

But the real win? The ads don't feel like ads. They feel like helpful contributions to conversations that are already happening. And that's only possible when you start with real customer insights, not marketing assumptions.

Your turn

If you're planning to test a new channel (or improve an existing one), start here:

This week: Schedule three customer conversations. Ask about their challenges, not your product

This month: Run an AI analysis on your existing customer feedback to spot patterns you've missed

This quarter: Create ad variants testing different insight themes, and track which ones drive not just clicks but engagement

And if you're curious about Reddit specifically - the platform is brilliant for reaching technical audiences who've tuned out traditional advertising. But only if you do the insight work first.

The hard truth? Most ads fail not because of poor targeting or weak creative execution, but because they're saying things that don't actually matter to the audience. Customer insights fix that.

Your customers have already told you what ads to create. Are you listening?

Talk soon,

Lucy 👋

P.S. We've documented our complete customer insights process in our Notion guide - from what questions to ask, to when to ask them, to how to turn them into action. Check it out here.

And if you want to chat through how to use customer insights to improve your ads (on any channel), just hit reply. I'm always up for talking about this stuff.