The team at Kooba have written an awful lot about AI in the past year.
With that in mind, I thought it was worth curating and condensing our thoughts, and offering a more comprehensive overview of AI adoption across our design and development teams. This should also (hopefully) serve as an insight into how AI is changing the web design industry more broadly.
Here’s 5 key ways Kooba have embraced AI, including links to more detailed articles on each topic:
1. Ideation and research
It may be obvious, but AI tools have made a lot of the “basics” easier for digital agencies like Kooba. This is especially true when it comes to discovery and research, a crucial phase in all of our projects.
Accelerated competitor analysis
Strategic design work relies on research. Lots and lots of research. By accelerating our ability to search for industry benchmarks, competitor strategies, and visual inspiration, tools such as ChatGPT and Claude save valuable time and allow for the collection and analysis of far more data than otherwise possible.
Read about the value of competitor analysis here: https://www.kooba.ie/journal/know-thy-enemy
Rapid prototyping
AI tools exist for a wide range of design and development tasks and allow for the rapid creation of wireframes, interfaces, and even websites. Whilst we still don’t trust the quality of these tools for actual client work, they are invaluable for creating quick mockups and prototypes during the early stages of a project. Rather than having to wait for help from developers, designers can see how a component may function using these AI tools, and can make adjustments accordingly.
Administrative work
UX design involves large volumes of data management and organisation, which must be done meticulously and methodically to ensure accurate results. By building our own internal GPTs within ChatGPT, our team can automate these otherwise repetitive tasks, and redirect time and energy towards more strategic and creative work.
2. LLM visibility and GEO
ChatGPT claims to receive over 1 billion prompts every day.
For obvious reasons, ranking highly in the responses to these queries is critical, since the most valuable, highest-intent users are often the ones asking LLMs about your industry or services.
Optimising your content for LLMs is known as generative engine optimisation (GEO), and when done correctly, GEO allows you to reach these high-value audience members directly.
Content GEO
Kooba now have almost a year of experience working on GEO for a range of clients. This means that our design, development and content teams understand the various ways in which content can be restructured and rewritten to rank higher within LLM responses and search engine AI-overviews.
To summarise our best advice: clear, well-structured, useful and contextual content will be rewarded by AI tools, and will drive high-intent users to your website as a result. This is now a standard consideration for Kooba designers when creating any content layouts.
Read our latest guide to GEO here: https://www.kooba.ie/journal/optimising-for-generative-search-how-content-structure-and-web-design-shape-geo-success
Technical GEO and schema
Just as with traditional SEO, there are crucial technical factors that contribute towards GEO. By building websites which are easily accessed by AI tools, and by making sure the content on each page is described semantically, we can increase the likelihood that our clients are mentioned by AI tools and AI overviews.
One important element of this involves schema. These are elements within a website’s HTML which indicate how content should be interpreted by non-human readers. These allow marketing teams to designate content as Q&As, FAQs, and other useful formats, which are in turn ranked higher by LLMs.
3. AI segmentation
This is the big one.
AI-segmentation promises to completely transform the nature of digital experiences and web design, allowing for highly personalised content to be displayed for each user.
Put simply, user data can be channeled to a LLM, which can then categorise users according to predefined personas, and provide an appropriate experience to each user accordingly.
Read an overview of AI segmentation here: https://www.kooba.ie/journal/the-rise-of-intelligent-websites-what-cmos-need-to-know
Segmentation for ABM
From a lead generation perspective, AI segmentation is an incredibly exciting tool. Account-based marketing (ABM) is based on the concept of hyper-personalised outbound messaging, but this can now be complemented by an equally personalised inbound funnel. Imagine, for instance, a member of your target audience being shown content specific to their industry and role when landing on your website.
See our guide to ABM here: https://www.kooba.ie/journal/the-missing-piece-in-most-abm-campaigns-your-website
Rules-based and generative segmentation
Kooba have already begun working on rules-based segmentation systems for different clients. This is a relatively simple process, as the personas used and experiences “served up” to each user are predefined and prebuilt. This means that Kooba can retain complete control over what the end user actually experiences, and ensure that the website remains accessible and usable.
An alternative approach is generative segmentation, in which AI tools build unique experiences for each individual user in real-time. This allows for even more personalisation, but comes with an as-yet intolerable level of risk. That said, this could be a transformative method for designing websites in the near future.
Read a full comparison between these two methods here: https://www.kooba.ie/journal/segmentation-vs-generation-choosing-the-right-ai-strategy-for-personalised-digital-experiences
4. Synthetic personas
Personas are a vital element of the UX process, serving as sample users who represent specific attributes, motivations, and behaviors. Typically, these personas are created based on insights from analytics data and interviews, but this information can also be used to build synthetic personas.
These are AI agents which we can use to test designs from the perspective of different user groups, allowing for the rapid validation of designs without the need for costly user testing campaigns.
Rapid testing and validation
The main benefit of synthetic personas (also known as synthetic users) is the increased flexibility and speed of testing. It is important to note that these tools cannot replace proper user testing but are instead suited to testing and validating solutions that would otherwise be neglected due to budgetary or time constraints.
A minor feature update, for example, could be validated instantaneously, despite being too small to justify human testing.
Long-term project savings
A major upside of synthetic personas is therefore the financial value they offer over the course of longer projects. By identifying risks early in the design process, they allow for fast and easy remediations, and minimise the time needed by clients to validate any new solutions. Inversely, findings from the project itself can be fed back into the synthetic personas, creating more accurate insights for future work.
Read all about how we use synthetic personas here: https://www.kooba.ie/journal/what-are-synthetic-ai-personas-and-how-can-they-improve-conversion-rates
5. Multi-lingual translation
AI has already proven to be transformative in the management of large website suites. For clients with international footprints, organising digital content across different languages and regions was previously a time-consuming process, and would limit the speed at which content could be shared.
Current AI tools allow for the automatic translation of content with the click of a button, meaning that a single centralised CMS can handle publishing across an entire suite of websites.
Multi-site management
AI content adaptation is of particular value to truly global brands. Our work with the WaterWipes website suite is a great example, with 18 regional sites spread across 9 different languages. By integrating AI-powered translation tools into a central CMS, Kooba have already dramatically improved the authoring experience for the WaterWipes content team.
Regional segmentation
Where this gets really exciting is with regional and cultural differences. Here we can use the AI-segmentation techniques described above to provide each user with regionally appropriate messaging and content, helping manage perceptions of a brand across diverse markets and audiences.
You can read more about AI and multi-site management here: https://www.kooba.ie/journal/why-localised-websites-are-the-new-global-standard
Conclusion: The beginning of the end, or the end of the beginning?
There has been no shortage of doomsday predictions regarding AI and web design, but the truth is that there is an awful lot to be optimistic about.
As we’ve learned in the past year, AI tools not only make it easier to do the work we’ve always done, but also allow us to deliver value to clients in entirely new ways. In reality, web design is at the cusp of a complete transformation, which will benefit both agencies and clients.
To learn more about how you can integrate the latest AI technologies into your digital solutions, just reach out to Kooba’s team today.
