Speaker 1 (00:00):
Today I'm going to give you an exact step-by-step playbook on how to roll out AI in your business, test it without burning through your budget and build a machine that gets smarter every single day. No hype, no fluff. Let's get into it. Most people running businesses right now are drowning in noise. Every week there is a new AI tool, a new platform update, a new thing that you are supposed to be doing, and somewhere in the middle of all that, the actual job of growing your brand, acquiring customers, and building something that lasts, it's getting harder to focus on. That's why I built this podcast, The Unblock. I'm Oliver Bruce. I'm the founder and CEO of Pinpoint Media, a performance marketing and paid media agency based in the UK. We run campaigns for some of the most ambitious brands in the world. We live in the data, we test the creative, and we operate at the intersection of where creativity meets performance.

(00:50):
And over the last few years, I've watched AI and automation completely reshape what's possible for marketers and business owners. Not in theory, but in practise. In the ad account, in the creative studio, in the way brands acquire and retain customers at scale. This is for anyone essentially running a business or a brand who knows things are changing fast and wants a way to stay ahead. Whether you're scaling a startup, leading a marketing team, or just trying to figure out what AI and automation actually means for your bottom line, the unlock is built for you. No hype, no theory, just what's working, what's changing, and what you should be doing about that. Now, let's get into the episode. Every single day, I speak to founders, business owners, and CMOs who are drowning in the noise of AI. They see the headlines, they buy the software, they slap an AI label on their processes, and then nothing happens.

(01:42):
They're kind of stuck in this groundhog day loop, if you will. They open up a new chat window, stare at a blank screen, type in a generic prompt, and then get a generic answer back and waste 20 minutes trying to fix it. And then they do the exact same thing the next day. It is actually mental, and it's a huge waste of time and a very dangerous mistake. Because what you're doing is you're treating AI like a search engine. You're treating it like a magic wand that is going to fix all your problems overnight, but AI has zero compounding value if you don't build a proper system around it. The game has changed completely. The businesses that are winning right now are not the ones that are using the most expensive tools. They're the ones that know how to roll out sort of tiny tests and iterate new processes.

(02:25):
Today, I'm going to give you an exact step-by-step playbook on how to roll out AI in your business, test it without burning through your budget and build a machine that gets smarter every single day. No hype, no fluff. Let's get into it. So the friction point, the reason most AI rollouts fail is because businesses try to do everything at once. They try to automate their entire sales funnel, their entire customer services department, and their entire content engine all at the same time. You can't just broad brushstroke it. That is a guaranteed way to fail. Frustrate your team and waste a hell of a lot of money. You have to start with what I call tiny tests. A tiny test essentially is about finding one specific, repeatable, high friction task in your business and isolating it. You're not trying to change the whole company. You're trying to build out one single bridge.

(03:11):
The tiny test step-by-step framework. Here is the exact process. Step one is to audit the friction. Look at your team's workflow and find the tasks that happen every single week. The ones that require a lot of heavy lifting, but do not necessarily require high level strategic thinking. Maybe it's taking an hour's long podcast and turning it into 50 short clips, a blog post and SEO transcriptions. Maybe it's summarising sales calls to extract objections, maybe even formatting weekly client reports. Find that one task, just that one task. Step two is map the manual process. Before you even touch AI, write down exactly how a human does the task right now. Step one, step two, step three, what data do they actually look at? What decisions do they make? What does a finished output actually feel like? If you can't map the manual process, you can't train the AI to do it.

(04:02):
It's kind of as simple as that. Step three is to build the baseline prompt. This is where most people get it wrong. They ask the AI to write a blog post. That is not necessarily a brief. That is just a wish. You need to give the AI context. You need to tell it what it is, what it needs to do, what the goal is, and what the rules are. Something, for instance, like this, you are the lead copywriter for Pinpoint Media, and your goal is to turn this podcast script into a 500-word blog. Do not use generic jargon. Use short, punchy sentences. Follow our brand guidelines. See attached. Context in, quality out every single time. Step four is to run the test in parallel. Do not replace the human yet. Have the human do the task. Have the AI do the task. Compare the two outputs side by side.

(04:48):
Where did the AI fail? Did it sound robotic? Did it miss a key piece of data? Did it actually get the wrong tone? This is the most critical part to the tiny test. You are not looking for perfection on day one. You're looking for a baseline that you can iterate on day after day. Hey guys, it's only me. I though this might be really useful for you. If like me, you found that one of the biggest headaches when running a business is actually managing money across different tools, different currencies, expenses, then I think I found the solution for you. InCard is built to essentially fix this. Now, InCard is a new financial platform designed specifically for high growth modern businesses. It gives you multicurrency accounts, connected banking, and smart spend management all in one place. Now, the really cool part is that with rewards, you get up to 2% cash back on things that you probably are already spending money on, like advertising maybe, SaaS tools, travel.

(05:43):
Every time that you guys are spending, you generate points and you can redeem them instantly for cash directly in the InCard platform. If you're building or scaling an online business and you want a smarter way to manage your finances, then check out InCard using the link in this description. Thanks for taking the time to listen to this. Now let's get back to the episode. So the iteration loop. Here is where most businesses stop. They run the tiny test. They see that the AI is not necessarily perfect, and then they kind of give up. This is the wrong move. Action is not enough. Feedback is required. If you just run the tiny test and stop, you will have zero compounding value. The magic happens when you feed the results back into your system. Let's say you're using a tool like Manus, for instance, to automate your market research or your copyrighting.

(06:30):
You do not just run the prompt and take the first output. You build a feedback loop. When the AI gives you something that works, you document exactly why it works. You note it down. You add your knowledge back into the bank. When it gives you something that doesn't work, you don't just fix the output manually and move on. You go back to the baseline prompt and you add a constraint. Never use this phrase in today's fast-paced digital world, for instance, or always open up with a question, or maybe keep every sentence under 20 words. You feed the failure back into the system and the system gets smarter every time. So building the knowledge bank. We, for instance, use clawed projects pretty religiously here. We upload our brand guidelines, our successful proposals, our failed campaigns, a transcript, for instance, from our best sales call. We give the AI the full context of our business.

(07:22):
So when, for instance, a new hire comes in, when we need to spin up a new campaign, we're not necessarily starting from zero. We're starting from a foundation of compounding knowledge. We are compressing months of onboarding into literally days. This is the difference between a tool and a system. A tool does one thing. A system gets better every time you use it. So scaling up, once you kind of validated the tiny test, let's say that you've nailed a process for versioning a podcast into shorts using Opus Clips and Manus, you don't just stop there. What you have to do is you synthesise it. You create an SOP, a standard operating procedure that says, "Here is the exact prompt. Here is the exact tool. Here is the exact input required." You make it so clear that anyone on your team can run without asking any questions.

(08:07):
Then, and only then do you move on to the next friction point. You build your AI infrastructure, essentially brick by brick. First, you automate content versioning. Then maybe you automate lead enrichment. Then you could potentially automate initial customer support triage, for instance. You are building a machine. And like any machine, it needs to be maintained. You need to consistently order the outputs. You need to consistently feed it with new data. You need to let the data dictate your next move. If for instance, you run a tiny test on a new AI workflow for outbound sales and the conversion rate drops, don't keep pushing it because it's cheaper and faster. You look at the data, you ask why? Was the, for instance, personalization tool too generic? Was the timing wrong? And then what you do is you iterate, you change, you do it differently. The difference that matters.

(08:56):
There's a big difference between businesses that use AI as a gimmick and businesses that actually use AI as a lever. The businesses that use it as a lever are trying to increase throughput. Essentially what they're doing is taking the best people and using AI to remove the heavy lifting so those people can focus on strategy, on relationships, on the things that actually move the needle. You're not trying to replace your team. You're trying to give your team superpowers. That's how I look at it. A real world example, for instance, let me show you exactly how this looks in practise for us. We had a massive friction point on our content distribution. We knew content is king, but distribution, as I've said before, is the kingdom. We need a volume, but volume alone is not the strategy. We needed high quality native content across TikTok, Instagram, and YouTube specifically for this podcast.

(09:43):
So we ran a tiny test. We took one 20-minute podcast recording. We fed it into AI to extract the best 60-second hooks. We then used another AI tool to format the captions. We managed to write the platform specific copy, one version for LinkedIn, one for Twitter, one for YouTube, et cetera. What we didn't do is just trust the AI blindly. We had a human review every single output. And what we found was that the AI was great at finding the hooks, but terrible at writing the LinkedIn copy. It was just too formal. It was too stiff. So we iterated. We went back to the prompt and we fed it the 10 best performing LinkedIn posts. We said, analyse the tone and the structure of these posts and apply the exact framework to the new copy. The next time that we ran the process, the output was 80% there.

(10:25):
The human only had to spend two minutes sort of tweaking it instead of 20 minutes writing it from scratch. We turned a process that used to take three hours into something that takes maybe 15 minutes. And we didn't necessarily sacrifice quality either. We actually improved it because we were sweating the asset, extracting the maximum possible value from that one 20-minute recording. That is the power of rolling out tiny tests and iterating. Guys, I think you might find this useful. Something that we've started to use in my businesses is InCard. It's a new financial platform designed specifically for high growth modern businesses. Now, if you're running a business, you've probably found that one of the biggest headaches is managing money across different tools, currencies, and expenses. So InCard essentially gives you multicurrency accounts, connected banking, and smart spend management all in one place. The best part, however, is that you can earn up to 2% cash back on everyday spend, such as ads, SaaS, and travel, earning you points every time you spend.

(11:27):
And you can redeem them instantly for real cash in the platform. Check out InCard using the link in the description. If you're building or scaling an online business and you want a smarter way to manage your finances, check out InCard. Let's get back to the episode. So your action plan for this week, here's what I want you to do. Not next month, this week. Number one is identify one single repeatable friction point in your business. Don't necessarily pick a massive strategic initiative. Pick a boring manual task that happens every week. Number two is to map the exact manual steps required to complete that task. Number three is to build a highly contextual baseline prompt. Give the AI an identity a goal and very, very strict rules. Number four, run that tiny test in parallel with your human process and then compare the results. Number five is to feed the failures back into the prompt, add constraints, and then build the knowledge bank from there.

(12:24):
This will compound, as I said, over time. Stop trying to make one massive AI transformation for your business. Start building a machine one tiny test at a time. Let the data dictate your next move. And remember, the goal is not to replace your thinking. The goal is to scale your execution. Do this consistently and you'll not just keep up with the changes in the market. You will dominate them. If you've enjoyed this and it's been useful, please subscribe below, follow the podcast and share it with someone else that you think might find it helpful. Thanks so much for listening or watching. If you're on YouTube, the latest episode of The Unlocked, remember to hit that follow, hit that subscribe button. Please share it with your friends, families, colleagues, and loved ones. This podcast doesn't grow without you. Honestly, it is really appreciated. I mentioned it earlier, but something that you may find super useful if you're a business managing multiple transactions across multiple platforms is InCard.

(13:19):
It's a new financial platform for modern online businesses giving you multicurrency accounts, connected banking, and smart spend management all in one place. Open a Euro GBP or USD account in minutes. Attach cards for expenses and earn up to 2% cash back on everyday spend like ads, SaaS, and travel. Check out InCard using the link in the description.