AI as a Thinking Partner

Out Of Home advertising has long been effective but hard to scale—until now. AdQuick makes it simple to plan, deploy, and measure campaigns with the same efficiency and insight you expect from online marketing tools.

Marketers agree: OOH is powerful for brand growth, driving new customers, and reinforcing messaging. AdQuick makes it easy, intuitive, and data-driven—so you can treat real-world campaigns like any other digital channel.

Learn more, visit AdQuick.com

When marketers talk about using AI, they almost always mean one thing: producing stuff. Writing the post, drafting the email, generating the ad variations, and making the content.

This is the obvious use. It is also the least valuable one, and it is the reason most marketing teams are getting a fraction of what AI could actually give them.

The highest-leverage use of AI in marketing is not production. It is thinking. Using AI as a strategic partner to research faster, understand customers deeper, pressure-test positioning, and analyse things you would never have had the time to analyse manually. This is the use that does not produce a visible deliverable, which is exactly why most teams skip it. There is nothing to show for it at the end of the hour. There is just a sharper decision.

This first edition of the three-part series is about that use. It is the foundation, because thinking badly and then producing fast just means you produce the wrong thing faster. Get the thinking right first.

Here is how to actually do it, concretely.

Most marketing teams are sitting on enormous amounts of unstructured customer information they have never analysed. Sales call transcripts. Support tickets. Customer interview recordings. Survey open-text responses. Reviews. Churn-survey comments. This data contains the most honest picture of your customer that exists anywhere, and almost none of it gets read systematically, because reading 200 support tickets and finding the pattern is exactly the kind of work humans avoid.

This is a perfect AI task. Not because AI is creative, but because it is tireless and consistent across large volumes of text.

The workflow. Collect the raw text. Export the last 90 days of support tickets, or 20 sales call transcripts, or all the open-text survey responses. Paste them into Claude or ChatGPT in batches. Then prompt with something specific:

"Here are 40 customer support tickets from the last quarter. Identify the recurring themes. For each theme, tell me how many tickets it appears in, give me two representative direct quotes, and tell me whether it points to a product problem, a messaging problem, or an expectations problem. Rank the themes by frequency."

What comes back is something your team has never had. A structured, evidence-backed map of what your customers are actually struggling with, in their own words, with the frequency attached. This is the raw material for better positioning, better content, and better product decisions, and it took 20 minutes instead of the three days of manual reading nobody was ever going to do.

Do this monthly. It is the single highest-value AI habit a marketing team can build.

Most positioning gets decided in a room by people who all already agree with each other. Then it goes live, and the market disagrees, and you find out six months and a lot of budget later.

AI is a genuinely useful tool for catching this earlier, if you use it as an adversary rather than a cheerleader.

The mistake is asking AI, "Is this good positioning?" It will be polite and tell you yes. The useful move is to explicitly instruct it to attack.

The workflow. Give it your positioning and then prompt:

"Here is our positioning statement and our three core messaging pillars. Act as a skeptical buyer in our target market who has seen a dozen competitors make similar claims. Tell me, specifically, where this sounds generic, where it makes a claim I would not believe without proof, and where it fails to tell me what is actually different about this product. Be blunt. Do not soften it."

Then run it again from a different angle:

"Now act as our toughest competitor. You are going to write the counter-pitch that wins the deal against this positioning. What is the weakness you attack?"

This does not replace your judgment. It does something more useful. It surfaces the obvious objections before the market does, while you still have time and budget to fix them. Most positioning is weak in ways the team cannot see because they are too close to it. AI, instructed to be hostile, sees those weaknesses immediately.

A real example. You are considering entering a new segment, launching in a new market, or going after a competitor's customer base. The right move is to research it properly. The reality is that proper research takes a week your team does not have, so the decision usually gets made on instinct.

AI collapses this. Not because it knows everything, but because it can structure and accelerate the research dramatically.

The workflow. Use an AI tool with web access, such as Claude, ChatGPT, or Perplexity, and prompt in layers. Start broad:

"I am evaluating whether to target [specific segment]. Give me an overview of this segment, their primary pain points, who currently serves them, and how those incumbents position themselves."

Then go narrower:

"Of the incumbents you listed, analyse the two largest. What are their pricing models, what do their customers praise in reviews, what do their customers complain about most often, and where is the gap a new entrant could attack?"

Then make it decision-useful:

"Based on all of this, give me the three strongest arguments for entering this segment and the three strongest arguments against. Tell me what single piece of information, if I could verify it, would most change the decision."

What you have at the end of an hour is not a finished strategy. It is a structured, evidence-informed briefing that would have taken an analyst several days to assemble, and it lets the team make the decision with real inputs instead of instinct. You still verify the critical facts yourself. But you start from a position of understanding rather than a blank page.

This is the most underused application and the hardest to describe, because it produces nothing you can put in a deck. But it is genuinely valuable.

Most marketing decisions get made with incomplete thinking because thinking properly is slow, and there is always a meeting. AI gives you a thinking partner that is available at 11 pm, has infinite patience, and will follow a line of reasoning as far as you want to take it.

The workflow is just a real conversation. You are deciding whether to cut a channel. Instead of deciding in your head in four minutes, you open Claude and think out loud:

"I am considering cutting our podcast. Here is what it costs, here is what the attribution dashboard shows, and here is what I suspect it actually does that the dashboard cannot see. Walk through this with me. What am I not considering? What would I regret? What is the cheapest way to test whether my suspicion is right before I make an irreversible cut?"

The AI is not making the decision. It is doing what a sharp colleague would do if you had one available every time you needed one. It catches the thing you skipped. It asks the question you were avoiding. It slows your thinking down enough to be good.

Marketers who build this habit make visibly better decisions, because their decisions have actually been thought through rather than rushed.

The thinking uses of AI are invisible, which is why they get skipped. Make one of them concrete this week.

Pick your richest source of unstructured customer data, support tickets, call transcripts, survey responses, whatever you have the most of. Pull the last 90 days. Run the pattern-analysis prompt from use one. Spend an hour with what comes back.

I am confident you will find at least one thing about your customers that your team did not know, or knew vaguely but had never seen with evidence and frequency attached. That one finding will be worth more than a week of AI-generated content.

This is the foundation. AI as a thinking partner makes every downstream decision sharper. Get this habit in place before you scale up production, because production built on shallow thinking just gets you to the wrong place faster.

Next edition, we move to production. How to use AI to actually create content and creative without producing the generic slop that audiences have now learned to detect and ignore. That is where most teams are getting it wrong, and there is a specific way to get it right.

See you at the next edition, Arindam