Our work doesn’t start with a single channel.

Clients come to us for all sorts of things: SEO, paid ads, web development, design content or any combination of them – but the real challenge is bigger than that.

We need to understand their competitive landscape:

  • What are their customers searching and how?
  • What is working for their competitors? Who is currently winning and who losing?
  • Where are the opportunities and which strategies are already saturated and table stakes?

This gives us the context for our plan… and is where Copilot researcher agent comes in.

It doesn’t replace strategic thinking and practical experience but it does let us gather, organise and synthesise information much faster freeing our team up to focus on its strengths, namely using our experience to scale their business and maximise their ROI.

How Research Has Changed

The research stage has always been a key part of marketing planning, but the way we do it has changed 100% as has the scale of it that we can carry out.

A detailed competitive review includes researching traditional search, paid ads and GEO as well as content gaps, website UX, trust signals and visibility beyond the web. For a full-service agency like ours, this matters a lot, because our research is the foundation everything we do.

Copilot has allowed us to scale our research and so develop campaigns more quickly and from a bigger base.

Learning To Use AI

AI tools come with a learning curve. Copilot is very approachable and so can fool you that you can skip the learning phase but knowing how to use it and knowing how to get the most out of it are completely different things. The 80:20 of Copilot is to focus on your skills not it’s settings.

Using clear constraints when writing prompts, asking for sources and assumptions to be explicit and verifying steps will lead to a step change in quality. When we moved across to Copilot we treated it as step change not a tweek or evolution. We tested several tools before setting on Copilot researcher, although clearly AI is evolving so quickly this might change.

Having done that we looked at team training and decided on an in-person training course with Acuity Training to upskill our team and combined ensure that everyone had the skills needed and was also using the same language to communicate about Copilot. We also made sure that we had a team Copilot specialist and made sure that they were in the office for the first few weeks to help out.

How We Use Copilot’s Researcher Agent

We use Copilot to help us process, synthesise and structure information – not replace the actual judgement needed to make sense of it all. When we use Researcher, we aren’t looking for a finished strategy. We are looking for a very wide initial view of a client’s competitive landscape, which it can produce so much quicker than manual research. We then use this to decide which areas to drill down into in more detail.

It helps us to pull a huge amount of raw data across multiple channels to build up an accurate picture of the competition. Our prompts ensure that Copilot is ‘marking its own homework’ as it goes but we will also often other AI’s to critique and double check this analysis before moving to the next step.

Why We’re Careful With It

AI-generated research comes in a neat package which can be very misleading. It is always polished and professional but can extrapolate inappropriately and is not good at pointing out uncertainties. This is especially important in niche and local markets where specific data and historic data can be poor or limited.

That’s why Researcher isn’t the whole process. Where recommendations are sensible, clearly well-reasoned and it provides relevant citations we will put them on a list for further work and review as we would work from a more junior member of our team. As ever the issue often comes down to where we can provide clients’ the biggest bang for their buck. Not all good ideas can be implemented in a world of limited time and resources.

Final Thoughts

Copilot’s Researcher function has become a key part of our process. In the early stages of a project we can get upto speed far more quickly for a client, it fits perfectly. We now gather and organise information faster, and build a much stronger starting point – all while giving our team more breathing room to focus on the work that creates real value.

It doesn’t replace strategy, specialist knowledge or experience, it makes them a bigger focus. The goal isn’t just to produce as much research as possible, but to produce better decisions from it.