The Conversion Playbook: 5 Experts on Turning Traffic Into Sales

15th September by Chidinma Itsuokor

For the first time since 2020, we gave a whole series over to conversion rates.

Not the narrow, tired version of that phrase. Not “run a few more A/B tests and hope.” The real thing: everything you can actually do to get more of the people already landing on your site to buy. Because conversion rate is the one lever that quietly improves everything else. Lift it, and every visit from every channel you already pay for- your SEO, your email, your ads, your social- starts working harder without another penny of spend.

So we sat down with five people who do this every single day. A brand strategist obsessed with trust. A former LinkedIn growth lead who now hunts for the mistakes killing ecommerce sales. A CRO consultant who spent two decades in the numbers, including a stint at Amazon. A site-speed specialist trusted by 750 companies. And a marketer building AI-native conversion tools.

They came at it from completely different angles. But listen to all five back-to-back, and something clicks. They are describing one system. Here is the whole masterclass, and how it fits together.

It Starts With Trust, Not Tactics

Here is the mistake almost everyone makes. They think conversion is a checkout problem. Fix the buttons, tidy the forms, shave a step off the funnel, done.

Rebecca Boyle, brand strategist and fractional CMO at Lissah & Boyle, opened the series by blowing that idea up. Her focus is the trust gap: anything in your messaging that stops a shopper doing what you want them to do. And it does not live at the checkout. It lives everywhere.

“It’s how you’re coming across generally as a brand. It could be on social media, in any marketing channel, on your homepage, throughout your website, on your product pages and through to the checkout. You don’t need to throw everything in, but if you add little things along the way at all those different stages, it can really help.”

The instinct, when someone says “trust,” is to slap some reviews on the page and move on. Rebecca’s point is that trust runs far deeper than a star rating. Yes, roughly 80% of shoppers check reviews first, because a stranger’s word carries more weight than yours. But the real lever is a brand that cannot be copied.

“People think the USP has to be something extraordinary. It doesn’t have to be. It can be a very small thing, but it’s just something that your business does differently.”

“We’re a family business” is not a USP, she warns, because someone can claim it tomorrow. A genuine, consistent personality is uniquely yours, and it is what lets people pick you on instinct when every product on the page looks the same. Her tool for holding that line is a tone-of-voice checklist, ten points maximum, so the founder, the marketer and the support team all sound like one brand. Her favourite test: is this how you would explain it to a friend in a cafe?

And when you do reach for proof, be specific. A vague “great product, love it” testimonial does nothing. A review that names the problem, the fix and the result builds real trust. Same with guarantees: link to a page that spells out exactly how it works, so “money-back guarantee” stops being a slogan and starts being a reason to buy.

The Mistakes Quietly Killing Your Sales

If Rebecca zoomed out, Ashu Dubey zoomed all the way in.

Ashu ran product growth at LinkedIn for five years before founding Alhena AI, and he has spent a long time staring at what stops people converting. His pitch was refreshingly blunt: it is easier, and more useful, to name the mistakes than to preach the ideals. He gave us five.

The first is making every CTA scream. That shiny new virtual try-on button feels exciting, but on a product page it is competing with Add to Cart, and attention is a zero-sum game.

“It’s basically an attention game on the product listing pages, and you want to be clear what you want the users to do on that screen. There is a principle: don’t make me think.”

The second is hiding your best sales associate behind a chat bubble. Modern AI agents can run a quiz, analyse a photo, build a routine, recommend the right product, and yet engagement on even the best chat widgets sits under 1%, because we have trained shoppers to only click that bubble for support.

“If less than 1% of your users are engaging with your best sales associate, who is hiding behind a chat bubble, you are doing a disservice to your users and to the technology you worked so hard to onboard.”

The fix is contextual entry points across the site. On a moisturiser page, offer “let us help you find the right moisturiser” right there, instead of burying it in the corner.

The third is assuming shoppers know what they want. Classic ecommerce dumps a grid and a wall of filters on people and walks away. A good store associate does the opposite: asks a couple of questions, then narrows things down. Ashu’s team nudges shoppers who have stalled, asks two or three questions, then presents the three best-fit products. The people who engage, he says, buy at four to five times the rate of those who don’t, with a healthy jump in order value too.

The fourth is forgetting that shopping is social. In real life we shop with friends, we send photos, we ask “does this work?” Online is dry by comparison. Make recommendations easy to share, Ashu argues, and you pull that social layer, and those friends, back onto your site instead of losing the conversation to WhatsApp.

The fifth is the sharp one: assuming shoppers understand what modern AI can even do.

“If you don’t educate users that these are the capabilities, they won’t ask those questions, they won’t get that help, and they’ll leave thinking it was a normal AI experience.”

Build the tool, surface it, and tell people what it can do. Miss any of the three and the whole thing is wasted.

Read the Story the Data Is Telling You

Carolina Dibb spent the best part of 20 years in ecommerce, including a stint at Amazon, before setting up her own CRO consultancy. Her guiding light reframes the entire discipline.

“For me, it’s all about the narrative, the story behind the numbers. It helps anybody working in that space to remember that there are humans on the other side of those numbers, and they’re behaving in certain ways for certain reasons.”

The trap she sees most is walking in with a preconceived idea and “fixing” whatever you personally dislike. Her advice is to let the data lead, because your gut is usually wrong about where the real problem sits. Start by remembering that a conversion rate is an output, not a starting point.

“My conversion rate’s low. First of all, that’s a relative metric, and it doesn’t live in isolation. It is an output of two other metrics, your traffic and your orders.”

So you go granular. Is a particular product type dragging the number down? A traffic channel? A specific page? Patterns only surface when you zoom in, and that is where a real hypothesis comes from, one you can then test.

Her sharpest line is the one that has haunted the industry for years:

“The metrics are an observation, but we need to draw out the insight.”

Knowing your rate is 3% means almost nothing until you understand why. Carolina gets to the why by layering her data: performance metrics like conversion, AOV and bounce tell you what is happening; heat maps, which she calls valid and badly underused, show where people scroll and where they skim straight past the thing you worked hardest on; and direct customer research adds the human context, something smaller brands are perfectly placed to gather because they sit so close to their customers.

The Invisible Conversion Killer

Then there is the problem nobody can see, and almost everyone has.

Jordy Scholing co-owns RUMvision, trusted by over 750 companies to watch site speed and Core Web Vitals. He opened with a stat that reframes speed as a revenue issue, not a technical one.

“Psychologically, it’s proven that if a website takes over three seconds to load for the first time, there’s a 50% chance they’ll go back to the search engine result page. It’s literally one swipe with your thumb to go back to Google and buy the same product at one of their 20 competitors.”

And it is not just the landing page. One hiccup anywhere in the flow, homepage to collection to product to checkout, can trigger a bounce, so the whole journey has to be fast.

Jordy gave the clearest plain-English tour of Core Web Vitals we have heard. Largest Contentful Paint is how long the biggest element takes to render, like the first product image on mobile. Interaction to Next Paint is how quickly the site responds when you click, think the spinner on Add to Cart. Cumulative Layout Shift is visual stability, that maddening jump when something loads late and shoves the page down. You will also spot Time to First Byte and First Contentful Paint, the moment the page stops being blank, in the tools.

The mistake most teams make is checking once at launch and ignoring it for a year. Synthetic tests like Google Lighthouse simulate a visit; field data from the Chrome User Experience Report, visible in Search Console, shows what real users actually experienced, but only on Chrome and only across the trailing 28 days. If you ship changes regularly, you want real user monitoring watching continuously and alerting you the moment things slip. Best of all, this is the rare fix that lifts your SEO and your conversion at the same time.

Where AI Wins, and Where It Must Not

To close the series, Elena Kostova, Head of Marketing at AI-native experience platform Uxify, tackled the topic the whole series had circled but never landed on: how AI actually helps you convert. Her framework is the perfect capstone, because it ties everything before it together.

Start small, and always start by measuring. Then sort every opportunity into two buckets: could this break the site, or could it make me money with little risk? High-risk changes get the AI-assisted treatment, with a human in the loop. Low-risk, high-clarity changes you can hand fully to the machine.

“Site speed, removing friction, personalization, optimizing images. Very few things can go wrong, and the knowledge to do it manually is quite vast. Those problems are very easily fixed with just letting AI do the job.”

But there is a hard line she will not cross, and it is the same instinct Rebecca opened the series with.

“Website microcopy, website branding, how your website is looking, those still have to sit with a real person. It’s a bit counterintuitive to give away that privilege to show your personality and your brand and just let AI do it for you.”

Because AI works from averages. And as Elena puts it, if everyone lets AI drive their brand, we would all end up beige. Shoppers can already spot an AI-generated site, an AI headline, AI imagery, at a glance, and in the long run the brands selling genuine experiences are the ones that win. AI can help enormously. But you still need a person in the driver’s seat, steering the brand.

Your 2026 Conversion Roadmap

Theory is useless without action. Here is how to actually put the series to work.

First, fix trust. Draft a ten-point tone-of-voice checklist and run one product page through it. Where are you asking people to trust you without giving them a reason? Add one trust signal, a specific testimonial, a clear guarantee link, at each stage of the journey rather than piling them all at the checkout.

Second, stop the leaks. Open your top product page and look at it cold. Count the CTAs competing with Add to Cart and cut the distractions. Find one place where you could guide a hesitating shopper instead of leaving them to the filters. And make sure people actually know what your on-site tools can do.

Third, follow the data. Break your headline conversion number down by product type, channel and landing page until a pattern jumps out. Add one heat map to your worst-performing page and watch where people really go before you decide what to fix.

Fourth, get fast. Run your key landing and product pages through PageSpeed Insights today. Anything over three seconds is your highest-impact fix this week, and it pays back on both search and conversion. If you ship changes often, set up continuous monitoring so you find out before your customers do.

Fifth, deploy AI wisely. List your CRO jobs in two columns, safe-to-automate and needs-a-human. Hand the first column to AI this quarter, starting with image optimisation and site-speed hygiene. Protect the second column, your brand and your voice, fiercely.

We take a week off between series so you can actually do this, because what is the point if you don’t take action? Pick one lever above and make one change this week. One change, done properly, beats five you never start.

The Through Line

Every expert we interviewed came from a different direction- trust, mistakes, data, speed, AI- and yet they kept arriving at the same handful of truths.

Trust, clarity, data, speed and smart automation. Pull all five, and you are not nudging one number. You are compounding the return on every visitor you already have.

For the first time in years, we spent five episodes on nothing but conversion. The tools have changed, the AI is new, but the through line is older than any of it: respect the customer, understand the customer, and make it easy for them to say yes.