Brilliant at the boring stuff: what AI genuinely does for our team.

The first two pieces on this hub were considered ones. Deliberately so.

AI adoption done badly is a real problem, and we wanted to say that clearly before saying anything else.

But here’s the other side of it, and it needs saying just as clearly. The last few months have been a quiet kind of productive. We use AI every single day, and it’s brilliant at most of what we throw at it. Not in a hype-cycle, everything-is-about-to-change way. In a Tuesday-morning, actual-real-work way. Tasks that used to take hours take minutes. Jobs that needed a specific person free at a specific time now just get done.

This is that list. Specific things, real time saved, no tool roundup. One rule of thumb we keep coming back to: AI is brilliant at work where being wrong is cheap, the task is repetitive, and a human can review the output in a few minutes. Everything below fits that description.

One thing to say before we start: AI output always gets reviewed by a human before it goes anywhere near a client or a channel. That’s not a caveat, it’s just how this works. The value isn’t in removing the human. It’s in what the human gets to spend their time on instead.

01

AI for competitor research and content planning

Competitor analysis used to mean opening what felt like hundreds of tabs, trying to absorb everything as if you were revising for an exam, and praying you’d covered it all by lunchtime. Now we give AI the brief, where do competitors rank, what content are they covering, where are the gaps in their messaging, and it comes back with a structured scan in minutes. We still read it with a critical eye. We still make the judgement call. But the two hours of tab archaeology? Gone.

Pillar content planning gave us a similar problem. Scoping out which content clusters to build, which keywords to prioritise, what angle each piece needs is strategic work, but to get to the starting point took the best part of a morning, at least. Now we use AI to map the basics first: the clusters, the keywords and the content briefs. Then we breathe and think. The thinking is so much faster when someone else has already done the legwork.

AI for understanding B2B tech sectors

Working in B2B tech marketing means some of our clients operate in genuinely technical sectors. AI has become the colleague we ask to explain something in plain English, then push back if we don’t follow, then check our paraphrase to make sure we’ve got the gist right. No awkward calls. No pretending we got it the first time. The right explanation, available the moment we need it.

02

Using AI for SEO, content optimisation and website copy

Checking content for SEO and AI search was one of those essential but unglamorous, time-consuming jobs nobody enjoyed once the writing was done. Are the keywords in the right places? Is the meta information correct? Is the article length right? Is that repetition creeping in after the fourth draft? We now run content through AI as a final check, making sure it’s working as hard as it should for search. It catches things we miss after staring at a draft too long. Ten minutes taken instead of an hour. The same job, done properly and at scale, is a big part of our B2B SEO work.

FAQ schema for SEO

This one surprised me personally. I’m not a developer, but I knew all about the importance of FAQ schema for search visibility. I needed to add it to a page, had no idea where to start, and used Claude to write the code, explain where it goes, and confirm it was working correctly. Start to finish: about forty minutes. A job that would otherwise have waited for the right window in someone’s diary.

Landing page and ad alignment

Landing page and ad alignment sounds simple and isn’t always. We know Google requires landing pages to match the ads that lead to them, but does the page actually deliver what the ad promises? Is the messaging consistent throughout? Is the call to action in the right place? We run this as a quick AI check before anything goes live. The thoroughness is impressive, and it’s caught things we might have missed. Getting that match right every time is the heart of good conversion and landing page work.

Streamlining website copy

You know the pages, the ones that have grown organically over several years, different people tweaking them over time, now taking four scrolls to get to the point. We’ve used AI to find where a page buries its own message and strip out what’s slowing it down. Not a replacement for a proper content audit, but a very useful place to start.

03

AI for content repurposing, email marketing and prospect comms

Repurposing long-form content is a favourite of ours. A whitepaper can become a series of social posts. A webinar becomes a blog summary. The structure and thinking already exist, AI just reshapes it for a different format. We still rewrite and edit, but the blank-page problem disappears.

AI for email subject line testing

If you give AI the audience context, the campaign objective and the tone, it will generate ten subject line options in the time it takes to write one or two. The hit rate isn’t perfect, but the speed of getting to a range of options is, and it’s really useful for A/B testing.

AI for warm prospect communications

Not the emails themselves, we still write those, because the voice matters. Where AI is useful is the cadence: how often to be in touch, what kind of content can get a reply, which assets are worth sharing at which stage. AI maps that plan out well when you give it proper context about the audience and what you’re trying to achieve.

04

AI for marketing admin: meeting notes, proofreading and more

Meeting notes and Google Meet summaries used to mean a recording you never quite got round to watching and precious time nobody had for writing up what happened. Now you can have transcripts turned into structured action points in a few minutes, with notes often better than the ones we used to write ourselves.

Conference notes follow the same logic. Screenshots of slides, quick summaries, action points. The kind of thing that used to take the morning after a full-day event. Now done before the train home.

Final proofreading catches things tired eyes miss, straightforward, but really useful at the end of a long day. Writing a short version of a message for a character-limited website notice, the kind of fiddly task that used to take twenty minutes of painful editing, now takes a couple.

Removing a face from a photo: we had an image we wanted to use for a web article but didn’t have consent for one of the people in it. AI turned it into a usable, generic image in minutes, a job that would previously have needed a designer, a brief and a turnaround time.

Regex. If you don’t know what it is, you don’t need to. If you do, you’ll understand why having AI write it on demand, whether that’s parsing data or working through a list of postcodes, is a minor miracle.