AI replacing CRO work - where does that leave us?

brandvoice

I love this space, but I'm hitting a wall and could use some outside perspective. We run a conversion rate optimisation agency - my brother handles the big picture stuff, I do all the client work. Clients stick around 12 months easily when it's a good fit, and our close rate after a call sits around 60-75%. So the offer isn't the problem.

Here's the kicker: we haven't brought in a new client in eight months. Existing ones are naturally cycling out, and you can feel the energy shift. Then there's the whole AI elephant in the room.

I've spent a lot of time thinking about what AI actually means for our niche. Manual coding? Dead in the last year. We let our two developers go and I handle everything now. Our real value was never the code though - it's understanding the intersection of marketing and website psychology, especially for e-commerce. But AI is creeping into that territory too.

Data analysis? Most tools now do that for you. Problem is, AI talks a ton of nonsense - maybe 40% garbage, 60% decent advice. Complex connections still throw it off completely. For smaller stores, under a million in gross sales, AI is probably enough. Above that, you still need a human brain connecting dots and building real structures.

Even if AI got better, ecom founders don't have time for this, or they want someone else to carry the responsibility. So I don't think AI kills our job entirely. It just makes finding clients harder - same story as always.

We're considering a restructure. How much AI do we bake into our service vs client work? Is it still a selling point to mention AI? that window might have closed already.

Curious what others make of this - especially the part about AI being a temporary overcorrection, where eventually business owners get tired of it and want a human pair of eyes again. or is that wishful thinking?

leadgen

Back in my customer success days, I learned the hard way that thinking beats just doing. Borrowed that from Rory Sutherland, but it rings true.

funnelhacker

Look into MCPs when AI gives you shit. Standard marketing agencies are incredibly replaceable. Smart operators use it to their advantage and stay ahead of the curve.

budgetburner

Everyone's losing their minds over AI in marketing. Yeah, it's a tool, but it's not replacing the strategic thinking that actually makes campaigns work. You still need someone who knows how to read a room, not just train a model.

And while you're worrying about tech shifts, don't forget the boring stuff-legal structure. LegalZoom's fine for sorting a LLC without paying a lawyer's mortgage. Just don't expect it to do your thinking for you. Adaptability starts with having your basics locked down, not chasing every shiny object.

metricsmuse

a 60‑75% close rate with 12 mon retention means the service and the offer aren't the problem. the gap is entirely in who's finding you before the call. the AI worry usually gets quieter once there's enough pipeline to see what's actually converting. when you look back at the last 3‑4 clients who signed and stayed the full 12 months, how did they originally hear about you? and what were they actively struggling with right before they reached out?

marketingmule

Honestly, that's the real secret - turning client results into a repeatable system. I've seen a few people in the beauty space try that with AI, but only the ones who had clean data and a solid framework actually made it work.

Out of curiosity, how do you handle organising the initial client data? spreadsheet hell or something more structured?

prpro

Sounds like the real issue here isn't the tool you're using, but more that the well has run dry. Eight months is a long time to rely on the same direction without refreshing the approach. Think of it like a garden-if the soil's been picked over, you need to plant new seeds elsewhere.

A couple of things that tend to convert faster than cold outreach at this stage: offering a free, focused audit to previous prospects (shows you remember them and value the relationship), or partnering with Shopify agencies who already have warm leads. Those doors usually open quicker than knocking on new ones.

ranktracker

I'm seeing exactly the same pattern in the technical SEO space. Log files don't lie - everyone's spinning their wheels running the same LLM experiments instead of deploying anything. Tool procurement has dried up across the board.

The shift you're describing makes sense. I've started doing something similar: instead of selling a one-off audit or a done-for-you optimisation, I'm now selling "AI-managed log analysis pipelines". The client keeps their favourite chatbot, but I'm the one building the knowledgebase - the site structure, the crawl priority rules, the indexed URL schemas - that the AI taps into for recommendations. That way my expertise becomes part of their operational loop rather than a one-time output.

It's not consultant versus AI. It's consultant as the architect of the AI's context. If you can structure your expertise so it's machine-readable and constantly updated, you become irreplaceable. That's the model that survives this downturn.

emailwiz

I've been in that exact spot - pipeline bone dry but closing two-thirds to three-quarters of the leads who actually got on a call. hurts, doesn't it? Means your offer's solid, but you're fishing in a puddle.

what shifted for me was getting brutally specific. Instead of "we specialise in nurture sequences" I'd find a business with a painfully obvious leak - say, a landing page with a 40% drop-off at the CTA. i'd screenshot it, slap a rough monthly value on the missed conversions, and send that directly to the person who could sign off. No big report, just one thing they couldn't unsee. Reply rates went from almost nothing to nearly one in five.

The AI fear is real, but honestly it punishes the generic more than the operator with a focused process. what does your daily outreach actually look like right now? Because if you've got that close rate, the fix is almost certainly in who you're talking to, not what you're saying

cpaoptimizer

Your close rate is solid, so the real issue is getting more calls booked in the first place. I went through a similar drought last year and started using Leadmatically to find Reddit threads where founders were already complaining about conversion problems.

It surfaces actual conversations instead of cold outreach, which helped me rebuild my pipeline without the usual spray-and-pray.

24fire

hit that same wall myself. When fulfilment eats up all your hours, growth work gets pushed aside. the turning point for me was getting brutally specific about who we actually helped best and why. instead of broad outreach, we started crafting messages around one core pain point we knew we could solve better than anyone else.

On the tooling side, I've been using something that watches real-time LinkedIn signals to spot high-intent leads. Keeps me from spraying and praying. Cuts the noise.

bouncekiller

Oh love, flashing the AI badge is the laziest flex in the room. Show me where your actual human call-making changes the cash flow. Frame it around decision quality, test priorities, and accountability - then let AI quietly trim your costs behind the scenes without becoming the main character.

paperclick

might sound like a dumb question, but are you actually selling CRO as CRO?

Because when it's done right for an ecomm store, it's way more than that. people think their deliverable is the offer - it's not. it's the promise.

you said you don't have an offer problem. I reckon you do

ppcnerd

Three clients this month-solid. But the real lesson here isn't the toolset or the cold outreach sequence. It's about momentum.

When you're trying to get a business off the ground, most people get stuck in the ideation phase. They build the perfect landing page, write the perfect script, and then wonder why nobody shows up. The truth is, client acquisition is less about the channel and more about the cadence. Keeping them engaged, showing them the impact before they've even paid you-that's where trust compounds.

I've seen this play out in video marketing too. You can have the best production value in the world, but if there's no follow-up mechanism, no loop that brings them back into the conversation, you're just broadcasting into the void. Same with cold outreach. It's a volume game, but only if you've got the system to hold attention.

The macro trend is this: attention spans are thinning, but trust is actually getting deeper with the right kind of repeated micro-interactions. People want to see proof before they commit. Show them the work, not the pitch.

sleepermode

A 60-75% close rate on calls suggests your offer isn't the issue-it's what happens before the call that's broken. You're not getting enough qualified conversations into that pipeline.

I'd stop thinking in terms of "we need clients" and start breaking the funnel into discrete stages: outreach attempts, reply rate, qualification pass rate, booking rate, show rate, close rate. One of those will be the weak link, and it's usually obvious once you lay the numbers out side by side.

Also, if one partner owns strategy and new business while the other handles fulfilment, that can work-but only if the entire sales process is documented to the point it doesn't live solely in one person's head. If your system depends on one partner's intuition, it's not scalable, and it's fragile. Get the qualification criteria, objection handling, and handoff steps defined in a shared document. Then automate what you can.

trafficcop

Oh honey, eight months with zero new clients isn't an AI apocalypse - it's a lead generation problem dressed up in a tech panic. agencies with close rates between 60 and 75 per cent and a year's worth of retention don't have a delivery issue, they've got a top-of-funnel crisis.

stop diagnosing the market and start fixing your pipeline. If you went eight months without closing anything, you either stopped doing outreach, your positioning got stale, or the channels that used to work just stopped. Pick one and fix it.

and please, drop the "we use AI" pitch. that train's already left the station. buyers don't care about your toolkit - they care about outcomes. lead with real client results and let the methodology whisper in the background.

The restructure can wait. go get your next five clients first.

dusting

Your close rate and retention are solid, so the real problem isn't AI coming for your job - it's that your pipeline has been on life support for eight months.

AI is a magnifying glass, not a replacement. It's just exposing a weak top of funnel. For stores doing seven figures plus, they don't actually want "AI" as a product - they want someone with a brain who owns the conversion problem and takes it off their plate. so instead of selling "we do CRO with AI", I'd frame it as "we make sure your traffic turns into cash, and we use AI under the hood where it actually helps".

If it were me, the next sixty days would be dead simple:

double down on outbound to stores doing 1M+ that are clearly scaling and seeing conversion drop. use your own split - like you said, "AI is 40% trash, 60% good" - and show them the 40% they're bleeding.
Build a few loud case studies and social content around "here's where AI alone gets it wrong and where human judgement added revenue". That positions you as "AI plus operator", not "AI competitor".
stop overthinking the restructure until you've booked more calls. The drought is telling you your acquisition engine needs a rebuild, not your core service.

Are you tracking daily or weekly numbers for outreach and booked calls right now, or is it more vibes-based? that alone would tell you where the real bottleneck lives

emailwiz

High close rate and strong retention? That tells me the offer's solid. The bottleneck is just getting in front of the right businesses before the call ever happens.

A couple of years back we hit a similar dry spell. Stopped all the broad outreach and got brutally specific about what a "ready" prospect actually looked like before we even sent a message. For CRO that means hunting for visible conversion issues you can spot from the outside - no social proof above the fold, CTAs buried below the scroll, checkout flows that look like a marathon. Things you can diagnose without touching their analytics.

The companies where the problem is obvious before the conversation are the ones that close fastest too. They already know something's off. You're just giving it a name.

What's your qualification process before outreach? How do you decide who makes the list?

weekendhustle

Oh, I've been there - drowning in the same tasks day after day 😅 One thing that genuinely saved my sanity was setting up automations for the really repetitive stuff. Even just a few simple ones can free up hours each week and let you focus on the actual business growth. Definitely worth exploring if you haven't yet?!