Camille Trent

Content & Community @ PeerSignal | Advisor @ Keyplay & AudiencePlus

I'm a pretty even split of Phoebe and Monica -- creative type who gets a little too competitive on the racquetball court. 

If you're into fancy bullets, I've got those as well:

✦ Content strategist with experience working in startups, as a lone wolf, and in an agency.

✦ Specialties: category content, analysis content, community, social media, copywriting, event & podcast strategy, content consulting

✦ Chill unless watching my Portland Trail Blazers blow another lead, then all bets are off. Dame Lillard for prez.

Bonus: I like numbers (math was my best subject growing up)

Member

Turn the internet's best ideas into your swipe file. Add notes and visually organize your favorite blogs, tweets, LinkedIn posts, podcasts, and videos.

I help growing businesses, from SaaS startups to service-based and B2B brands, build and scale marketing systems that drive predictable revenue growth.

Oli Bridge

I shipped what I thought was our best homepage in years for Bonjoro. It tanked signups by 50%. 

Here's what I got wrong, and why stats in your hero might be killing your conversion too.

The change:

We added performance stats (68% open rate, 49% watch rate, 23% reply rate) to the hero section.

It seemed obvious:

Show proof early → build trust faster → more signups.

What actually happened:

Engagement time halved, 26s → 11s
Signups collapsed.

The stats didn't build trust. They short-circuited the entire journey.

Here's why:

When someone lands on your homepage, their brain is trying to answer one thing first:

“What is this, and is it for me?”

Not: “Does it work?”

Stats answer the wrong question too early.

Before the stats, the journey was:

→ Land on page 
→ See product visual (video messages) 
→ Scan headline 
→ Understand what it is 
→ Read how it works 
→ Consider signup

~26 seconds of exploration.

After adding stats to the hero:

→ Land on page 
→ See product visual (video messages) 
→ Scan headline 
→ See stats 
→ Too much to process / added uncertainty
→ Leave

~11 seconds. No exploration. No signup.

The psychological shift:

Without stats, Bonjoro felt like: "Personal outreach. Human connection."

With stats in the hero, it felt like: "Conversion optimization tool with 23% reply rates."

Even though these are incredible results, visitors started thinking about effort vs. return:

"I have to record a personal video for every single lead?"

The stats made the work visible before the value felt real. And humans instinctively look for shortcuts, not more work.

What I learned:

-> Stats are powerful - but they belong after understanding.

-> You can't summarize the value prop with numbers and expect people to convert.

-> They need to understand before they believe.

-> Stats without context create doubt, not trust.

The fix:

We reverted the hero to the product visual.

Engagement time bounced back within 3 days.

The lesson:

Your homepage isn't a pitch deck.

Show the product first. Prove it works second.

Growth Marketing Consultant for B2B SaaS | Helping SaaS Founders Build Predictable Revenue

Most founders don't have a traffic problem. 

They have a buying problem.

I call this a "buying model mismatch".

The first thing I do with almost every client is map how their customers *actually* buy, not how the founder assumes they buy.

There are (generally) four buying modes:

→ Problem-based: "Something's broken, I need to fix it" 
→ Category-based: "I need a tool that does X" 
→ Competitor-based: "My current solution isn't good enough" 
→ Brand-based: "I want this specific product"

Most businesses are only truly optimised for one or two, not all four.

For a B2B ecommerce client last quarter, I ran a full GTM diagnostic in 4 days:

→ Audited their entire product catalogue
→ Mapped their competitive landscape
→ Identified the buying mode gaps

The fix looks different every time - in this case it was restructuring their website architecture and product taxonomy to match how their customers actually think and search.

Six weeks later:
↑ 39% sales growth year-on-year
↑ Best February in company history
↑ Orders coming in like clockwork

Predictable revenue doesn't come from working harder on the wrong model.

It comes from building the right system first.

If you're struggling to grow despite decent traffic, there's a good chance it's not a traffic problem.

It's a buying model mismatch.

So here's the real question:

Which buying mode drives most of your revenue, and which ones are you missing?

Every founder has an “oh shit” moment.

I found ours while looking back through some old Slack messages this week.

It was our CTO posting that Typeform had signed up to Bonjoro.

At first we were excited.

Typeform! Using our product!

Then a slightly uncomfortable thought crept in…

"Why would Typeform need Bonjoro?"

They were largely self-serve. Most users never spoke to a human.

So the signup felt... strange.

And suddenly the possibility hit us:

They weren’t a prospect.

They were researching the space.

A few months later they launched their own version of Bonjoro. 

And honestly, when it came out, we were a bit worried. It looked *really* good.

Super high UX design bar, as you'd expect from Typeform.

They even used some of the same language and ideas we’d been using, including a term I coined myself: “video funnels”.

For a moment it felt like we might get flattened. As in, fully wiped out.

But then something interesting happened.

Nothing.

Because despite how similar the products looked on paper, we weren’t actually solving the same problem.

Different ICPs. Different use cases. Different reasons people were buying.

They were building a video marketing tool.

We were building a personal video tool for sparking real, human connections with leads and customers.

From the outside, the products looked close.

In reality, the lanes were completely different.

We kept growing.

They quietly moved on.

A good lesson for founders:

Competitors often look scarier than they really are.

Pick your lane. 

Serve it well.

The rest is mostly noise.

Everyone's obsessed with LLM optimisation. But AI visibility is probably more of a positioning game.

I ran an AI visibility analysis on Bonjoro last month. Across 800 simulated AI conversations (ChatGPT, Claude, Gemini and more), Bonjoro shows up in 51% of responses for "personalised video messaging."

That puts us at #3 in the category, behind Loom and Vidyard.

Those companies have teams and funding orders of magnitude larger than ours. Yet we still show up alongside them in AI-generated answers.

Here's why I think that is:

We owned our category. We never tried to be everything to everyone. We'd been "personalised video messaging for SMBs" for years, consistently, clearly, relentlessly.

AI models have simply learned that over time. Which reinforces something I've believed for a long time:

Positioning is still the most important part of any marketing strategy.

TLDR; stop obsessing over LLM optimisation. 

Dominate your category. The AI will follow.

Important, so worth reposting:

"When ChatGPT or Google recommends a brand in an AI answer, that recommendation comes from the model's base knowledge, patterns baked in during training across billions of documents.

The citations you see at the bottom? Those are surfaced after the answer is generated, not before. It's post-hoc rationalization. The model didn't choose your brand because it found your URL. It generated an answer based on what it already knows, then pointed to sources that support it."

Leading SEO & AI Search at Semrush | Helping brands turn generate revenue across Google + AI answers

A lot of AI citation research is being misread right now, and it's costing marketers real strategic clarity.

You've probably seen studies flooding your feed: "We analyzed 50,000 AI citations to figure out what's working." "What 10,000 AI responses reveals about how LLMs choose brand sources."

The implicit argument in almost all of them is that the citations you see in AI answers explain why certain brands appear in those answers. Get cited more, show up more.

That's not how it works.

When ChatGPT or Google recommends a brand in an AI answer, that recommendation comes from the model's base knowledge, patterns baked in during training across billions of documents.

The citations you see at the bottom? Those are surfaced after the answer is generated, not before. It's post-hoc rationalization. The model didn't choose your brand because it found your URL. It generated an answer based on what it already knows, then pointed to sources that support it.

This matters because you literally cannot reverse engineer it. If you ask an AI where it learned something, it can't tell you. The model is too complex. 

Training data and cited sources are fundamentally different things, and no citation frequency study can reliably tell you which one is doing the work.

The one real exception: in cases where a model is heavily relying on RAG and the information is recent or limited, the cited sources can have direct influence on the answer.

Lily Ray has run experiments where she got AI Overviews to answer a specific question based on something published an hour earlier. That's real, and it's worth understanding. But it's a specific condition, not the general case. 

I've also seen Rand Fishkin bring this same concept up recently. And people just. keep. missing the point.

Citations aren't worthless data. Being cited frequently does correlate with showing up in AI answers. But correlation is not a strategy. 

Before you build a roadmap around citation tracking, make sure you actually know what you're measuring.

taplio

The Battle

Camille Trent

Oli Bridge

Followers

Posts

Engagements

Likes

Performance ScoreOli Bridge's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Camille Trent

Oli Bridge

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Oli Bridge's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreOli Bridge's LinkedIn Performance