"Let's choose an AI tool" – why that's the wrong first step
Many companies begin their AI journey by asking: Which tool should we use? ChatGPT? Copilot? Our own model?
The problem is that if you start with a tool, you’re thinking in terms of the tool rather than the problem. And that usually leads to expensive demos that nobody can really scale.
The better approach: start by asking what kind of problem you want to solve. Because AI isn’t universal; it excels at certain types of problems but is less suited to others.
In this article, we’ll show you the seven types of problems where AI has proven to be particularly effective, and how your business can get started with AI in a systematic way, rather than getting bogged down in frameworks.
What all these types of problems have in common is that they can all be implemented using a single platform, namely Make (make.com).
Make is a no-code automation platform that visually connects AI workflows: from data input through AI processing to output in your CRM, Slack, or any other tool.
Why most AI projects fail before they even get off the ground
We see that in many companies, AI initiatives fall short of expectations. The most common reason: a mismatch between the problem and the technology. This doesn’t mean that AI doesn’t work. It means that the wrong questions were asked. Instead of “What can AI do?”, you should ask: “Which problems in my company are, by their very nature, AI problems?”
This is precisely where a framework comes in that has been developed in practice with hundreds of teams: the 7 natural types of AI problems.
The 7 types of problems where using AI really makes sense
These seven categories cover the majority of real-world use cases across various sectors, ranging from marketing and operations to sales.
Summary
AI summarises large amounts of unstructured information – automatically, consistently and in real time. A typical example: all monday.com service tickets from a given week are automatically summarised and sent as a Slack report. This saves hours of manual work and improves transparency within the team.
Other use cases
Meeting minutes, customer feedback analyses, competitor monitoring reports and weekly status updates can also be automatically summarised using AI and forwarded to the relevant departments, individuals or tools. This saves a huge amount of time.
Text Creation
AI-generated content has a bad reputation – and rightly so, if it’s used without a strategy. Used correctly, AI can produce high-quality, brand-aligned content in seconds. The key lies in prompting and human quality control. Today’s AI supports not only text, but also images, video and audio. So whilst you cannot hand over the entire creation process to AI, you can optimise it to generate text more quickly and easily.
Where you can use it straight away
For blog drafts, social media posts, product descriptions, email campaigns and many other text-based tasks.
Klassifikation (Classification)
You have a lot of data coming in via monday.com, Asana or another project management tool –> enquiries, tickets, comments, applications – and you need to categorise it. AI does this in milliseconds, even with unstructured data.
Where you can use it straight away
Ideal use case for support ticket routing, lead qualification, content tagging and CV screening.
Sentiment Analysis
Do you want to know how customers, applicants or users react to your content? AI detects moods, intentions and emotions in text – across all channels. This provides you with insights that would never be scalable or analysable manually.
Where you can use it straight away
Customer reviews, social media monitoring, campaign feedback, and email correspondence with potential customers.
Data Transformation
Getting data into the right format used to be a tedious and error-prone process. AI can reliably convert unstructured or semi-structured data into structured outputs – and do so in no time at all!
A classic example
Long blog articles are automatically converted into social media posts in various formats. Content repurposing, data migration, form processing, legacy system integration and many other possibilities.
Search & Enrichment
AI-powered data enrichment significantly outperforms traditional enrichment tools – in terms of quality, speed and scalability. You can automatically retrieve and analyse information from the internet and integrate it into your systems to make better decisions.
Where you can use it straight away
Lead generation in CRM, market research, competitive analysis, supplier research.
Data Extraction
Extracting relevant data from emails, meeting notes or documents and transferring it directly to the CRM or other systems – this is one of the most powerful use cases for go-to-market teams. No more manual copy-and-pasting, no more forgotten follow-ups.
Where you can use it straight away
Email-to-CRM workflows, contract data extraction, meeting-to-task automation.
How to get started – the 3-step guide
Theory is all well and good. But how do you actually get started as a business? Here’s a practical approach that works in real life:
Step 1: Problem before tool – identify a specific, recurring problem in your business. Ask yourself: Does it fall into one of the seven categories listed above? If so, it’s a natural candidate for AI and you can use it to optimise your processes.
Step 2: Start small – choose a single use case. Not ten. Build it from scratch, test it, optimise it, and only then move on to the next one.
Step 3: Scale the pattern – as you work through Step 1, you’ll often spot further use cases that build on it. This creates a compounding effect: each new step improves the previous ones.
The most common mistakes when getting started with AI and how to avoid them
- Too many tools at once: Focus trumps variety. Choose one platform and master it before adding the next.
- No clear measure of success: Define what success means before you start. Time savings? Fewer errors? Revenue growth?
- AI as a replacement, not an enhancer: AI is at its strongest when it supports human decisions – not when it replaces them. Keep people in the loop.
- Lack of a change management strategy: Your team needs to be on board. AI projects often fail not because of the technology, but because of people.
Ready to deploy AI at scale in your business?
We help SMEs, marketing teams and executives not just to try out AI automation, but to integrate it into their processes on a permanent basis.
.svg)