When you start building a business online, there are plenty of services you can pay for.
Reports, dashboards, recommendations. Everyone has something that promises to help you understand what is happening and what you should do next.
But there is a place we can start for free.
Google Search Console and the standard version of Google Analytics give us a foundation for understanding how people find our website and what happens when they arrive.
They may not give us the most convenient experience or an answer to every question. But it is still Google. :) If we want to understand our presence in Google Search, its own information is a useful starting point.
Before deciding what else we need, we can get familiar with what is already available.
Why?
One word. Context.
We want AI to help us. It needs information, like you, me or anyone.
Making the right decision about what comes next requires information. You can filter it yourself by opening a report, looking at the numbers and then asking AI to do something with it.
But we can’t be everywhere all the time. My recommendation is always to bring AI in to save our attention for where it is required. Google’s information is a MUST for having a conversation about a website.
If you have a website, this is a no-brainer. Connect it with AI. That way, you will have more information, with less time, attention and focus required to build a meaningful conversation about the website with AI.
For me, Google is like lights on, lights off. It represents that kind of basic.
What will the connection be able to do?
Google Search Console and Google Analytics provide different parts of the picture.
Search Console helps us understand how the website appears in Google Search, including:
- Search terms associated with our website.
- How often our pages appear in results.
- How often people click.
- Which pages receive that visibility.
Google Analytics helps us understand recorded activity on the website, including:
- Where visitors come from.
- Which pages they visit.
- How they engage.
- Actions such as enquiries or subscriptions, when the relevant tracking is configured.
Our connection gives Alex read-only access to reports from both services.
We can ask questions, compare periods and investigate changes within our conversation. It does not give Alex permission to change the website or our Google settings.
The available reports depend on what the connector supports. Access to the information also does not mean AI automatically knows our business goals. We still need to provide that context.
How?
We built a custom connection using Google Cloud APIs and authentication.
These are the main steps we followed. The exact setup depends on the connector you use.
- Get the Google accounts ready. Make sure the website is verified in Google Search Console and Google Analytics 4 is collecting data.
- Decide what the connection should access. We wanted search and website performance reports, with read-only access.
- Set up a Google Cloud project. This is where we managed access to the Google APIs.
- Enable the relevant APIs. We used the Google Search Console API and Google Analytics Data API.
- Configure authentication. Set up the consent configuration and credentials required by the integration. The Google account authorising access needs permission to view the relevant website properties.
- Connect through a compatible integration. Enabling Google’s APIs and making their information available inside ChatGPT are two parts of the setup. A compatible connector or custom integration brings them together.
- Test both services. Ask for a simple report from each source using a clear date range. Compare the results with the corresponding Google report.
There were technical steps, permissions and things that did not work the first time.
But we got the connection working. :)
Now what?
There are two ways to use this integration that we keep coming back to: conversation and scheduling. You can use both.
Conversation
Bring a question into the conversation when you need it.
You may be deciding what to write next, checking what happened after publishing an article or trying to understand a change in traffic.
AI can look at the available information with you, compare it with your goals and help you explore what to do next.
The information becomes part of the conversation while you are working.
Schedule
When you know what you want to follow, you can set up a recurring review, where scheduled access to the connector is supported.
Perhaps you want a short update before planning your week. Perhaps you want to know how your chosen KPIs are progressing at the end of each month.
Decide what the review should include, what it should compare and what deserves your attention. Set up the scheduled task separately from the connection.
That gives the repeated work a place in your routine without you having to remember to ask every time.
What is the thought process behind it?
KPI.
The invisible connection between having this connector and the thought process supporting the best use of Google Search Console and Google Analytics.
KPIs are how we keep a conversation with our AI directed and efficient.
When we define our KPIs, they represent what is important to us and what success will look like.
Google Search Console gives us basic information about what people find us for and how they reach the website. Analytics shows what visitors are doing on our website and where they came from.
Thoughts are not easy things to get people or AI behind and working towards a shared goal.
KPIs are how we do it.
In this case, we need to think about questions like: What do we want to be known for? What is our vision? Our mission? What value are we creating? What need are we solving?
Those are just examples. The questions change depending on our product, stage of business, service and more.
But they are all thoughts. Thoughts only you, as the human, can answer. And KPIs are the best way to get anyone, including AI, aligned and working towards a shared goal.
Things to look out for
The correct website and property
Check that the connection is reading the intended Search Console and Google Analytics properties, especially if your account has access to several websites.
The period being reviewed
Ask which dates the report covers and whether the comparison uses equivalent periods.
A partial week compared with a complete week can create a misleading impression.
What the numbers mean
An impression is not a visit. A visit is not an enquiry.
Search Console clicks and Google Analytics sessions are measured differently, so their totals will not necessarily match.
What is actually being collected
If you want to understand subscriptions, enquiries or purchases, check that those actions are being recorded.
A connection can retrieve available information. It cannot fill in tracking that was never configured.
What the information shows and what AI concludes
A report can show that something changed. It may not explain why.
Ask AI to distinguish what it found in the report from its possible explanations. Then explore those explanations with the context you have about the business.
Optimisation
When it comes to marketing, optimisation is the key to everything.
This will keep changing and evolving as the KPIs change with the business.
Google’s reports give us an ongoing view of how people find and use our website, although the information is not all available in real time.
The world changes by the day. What people search for changes. How they find information changes. So what we track and ask AI to look out for needs to adapt too.
Review whether your KPIs still represent what you are trying to achieve.
If your priority changes from getting known to generating enquiries, the review you ask AI to prepare should reflect that.
Keep the measures consistent enough to learn from them. When you change them, explain why and make the new direction clear.
The connection can stay the same while the thinking guiding its use evolves.
Personal message
My message here is really about KPIs.
When I was a CEO and wanted to make sure we were all aligned, numbers were what I kept everyone connected to.
That is what I suggest here.
Numbers give us a neutral, shared reference point. Once we agree on what they mean and how we measure them, they are easy to track without too much explanation.
What I want to bring to light here is that KPIs are a thought process materialised in numbers.
So when you make the connection, the numbers will represent your thought process.
Here, it is easy to see how the thought process is as important as the technical side of implementing it. :)