Is AI ready to replace CRM? Four key considerations for modern organisations

Companies are rethinking data, security, and the adoption of artificial intelligence (AI) in customer engagement, and it’s playing out as a debate between AI and customer relationship management (CRM). We’ll explore the different approaches and look at four key issues you need to consider.

The rise of AI in customer engagement

With the massive momentum building around AI, and its potential use across marketing, sales, and aftersales processes to improve, or even revolutionise, how companies engage with their customers, it was only a matter of time before it started making headlines and industry “battlelines” were being drawn.

On the one hand, you have companies such as Klarna, who have announced they’re going to completely stop using a CRM altogether and replace it with pure AI usage. On the other, you have enterprise technology leaders challenging how this is even possible and that using their embedded solutions is the better, if not only, way to go.

As proof, you only have to look at the dramatic increase in AI investment of companies like Microsoft, who spent $14 billion in early 2024 alone. Meanwhile, Salesforce, through its Salesforce Ventures fund, doubled its investment to $1 billion over 18 months, with their AI tool Agentforce taking centre stage at Dreamforce. HubSpot has also ramped up, investing in Jasper in 2022, Caption in July 2024, and launching their own AI, Breeze, just recently. It's clear to see that these organisations quite literally have a vested interest in promoting AI; they simply have to make these investments profitable, and embedded AI is another reason for you to license their technologies.

But how much substance is there to this? Let’s take an objective look.

Data landscape

Something that should be fundamental to your AI strategy is where the relevant data currently resides. This becomes even more pertinent when exploring whether to buy into the “CRM is best” argument or look into something more tailored.

Whilst it is possible to leverage data sources that sit outside of these enterprise tools when using their embedded AI offerings, to achieve the speed and level of value they’re advertising really does require you to assume the approach of moving or copying your data into one of the many products they offer.

This could mean customising the CRM tool itself, storing it in a marketing automation tool, or even licensing and using a customer data platform (CDP). However, there are pros and cons to weigh with each approach. Bringing all that data into the platform doesn’t come with the sudden ability to do all the things you probably have on your AI wish list either—something we’ll touch on later in this blog post.

If the thought of committing that much of your data, and potentially business itself, to a single provider scares you, then whilst you may have a more difficult or even expensive time setting things up, the ultimate cost and value could be considerably lower.

When we look at other independent AI solutions, many offer out-of-the-box (OOTB) feature integrations with a range of enterprise technologies. These solutions can even integrate with cloud storage providers like Azure, AWS, GCP, and Snowflake, as well as what some of these cloud providers themselves offer. This makes it easier to understand why companies like Klarna are choosing to keep their data outside of big enterprise CRM platforms and still access the information they need, such as:

Rather than logging into their CRM prior to meeting a client, they can have AI tell them all they need to know about that customer and what to expect.

Rather than logging into a CDP to analyse the data and recommend audience segments, they can explain what the campaign is about using natural language processing (NLP). AI will also provide them with an audience list based on those it calculates it’s most relevant to from their own data sources.

In essence, this approach could offer the quickest path to the specific, and therefore most valuable, answer your business needs to make an informed decision.

Data security

Another major talking point around AI is the security aspect. If I use AI, will my data become visible to people it shouldn’t be? How can I be sure?

Companies such as those listed above, who are embedding it within the very fibre of their platforms, market this as a reason it can be trusted. As usual, the reality isn’t quite that simple.

When it comes to data governance and security, getting the use of AI “signed off” within a company doesn’t simply disappear because it’s embedded within the platform. Rather, most companies have had CRM-related technologies signed off for specific use cases.

Some common examples would be:

I’m putting my customer contact and purchase data into a CRM so my salespeople can manage the sales process, or;

I’m putting my customer preference data into a marketing automation tool so I can personalise my marketing content.

To take full advantage of AI, it needs as much data as it can have access to. This means your signed off use cases and reasons for those are likely to expand dramatically as you pull more and more data into these platforms—think personal data such as financials, logistical, or even medical.

Gone are the days when you could simply say, “I want to just put all this data in as it’ll help us.” Nowadays, it’s not just about what data you want, but also why, how long will it be kept for, and who will interact with it. As a result, any additional data needs will rightly have an expected level of due diligence, and it’s not going to be as simple as, “You already use our platform, you can trust us.” Over time, however, trust in these platforms is likely to grow as legal and data governance teams become more familiar with how data is stored. With fewer data locations to manage and more targeted data types tailored to specific AI use cases, oversight becomes simpler and more focused.

Speed of adoption

With all the noise surrounding AI, an important challenge lots of business leaders are faced with is how quickly they can extract some value from AI.

How quickly you make AI relevant, and therefore actually useful to your business, is arguably the one area where it might make sense to listen to some of the marketing messages coming out of events such as Dreamforce, Inbound, and undoubtedly Ignite later in the year.

With decades of experience, these enterprise companies can rightly state they have a huge advantage in understanding what sort of use cases their customers are likely to have across marketing, sales, and aftersales-related processes.

This knowledge has enabled them to develop templated “agents” (the modern day “bots” that some of us may remember) which can be deployed with little-to-no effort, effectively enabling “no-code” AI deployments to become a real possibility.

Now it’s true that these use cases might appear to be basic in nature, but that shouldn’t detract from how useful they are as a time, and therefore cost, saving activity if companies are prepared to put their trust in them. For example:

Having a marketing campaign “agent” that can help identify the most valuable audience based on data, or what content is most likely to succeed depending on specific customer needs or wider industry trends.

Having a sales “agent” that can take away all the hassle of recording activities, booking meetings, and even helping other real-life salespeople prepare for pitches.

Having a customer service “agent” that has access to all the necessary data it needs to be replying to customer queries 24/7, and every time it learns whether its responses were helpful or need to be improved.

All these things can be built outside of these enterprise platforms of course, but being able to access and use templates built by experts with decades of history and millions of customers to draw inspiration from will undoubtedly provide accelerated time to value with specific AI use cases.

What it won’t help you with, at least not without significant investment in a data strategy and bringing in the right talent to harness AI fully, is the “ask and you shall receive” future that a lot of leaders are thinking about when they talk about AI.

This utopia could be made easier by having an alternative AI solution running over the top of your own data sources, rather than looking to use embedded AI agents within your CRM platform.

Accessibility

Finally, AI used to be something that sat firmly in the ‘highly technical’ user domain, and to a certain extent it still is. However, with the rise of AI platforms such as OpenAI, DataRobot, and TensorFlow (other AI platforms are available), enterprise CRM companies have recognised they need to provide a more user-friendly and accessible approach to their offerings.

So in this area we find that even CRM powerhouses aren’t claiming to have an edge over those open-source platforms. Rather, they are addressing the gap so they’re not left behind. For example, within each of Microsoft’s, Salesforce’s, and HubSpot’s AI marketing, you’ll find references to how easy it is to implement and customise their templates, or even develop your own “agents” using their models; think low-code, drag and drop-style AI implementation for meeting any specific needs you may have, and learning as it performs those tasks to continuously improve.

Weighing your needs

In summary, the real answer when you’re asking the question, “Do I invest in an enterprise CRM-based solution, or do I look for alternatives?” is dependent on where you are on your journey and the value you are looking for.

If your organisation already has substantial data within a specific enterprise platform, there may be a faster way to achieve value. You could save time or even improve the quality of certain marketing, sales, or aftersales administrative tasks. This quicker path could be found by exploring the AI use cases already embedded within the platform itself.

If the data that will deliver the value you need is spread across multiple sources, consider centralising it in a single cloud storage location. If your data is already consolidated in one place, investing in an AI solution that connects easily to various sources—or even building your own—might be the next step. This approach could free you from the rising licensing costs of enterprise-level CRM platforms, while still giving your teams the insights they need instantly, at the touch of a button (or the tip of their tongue!).

Whatever your situation, whether one of these scenarios resonates or you’d like guidance on which might suit you best, Credera has MarTech and AI experts who are here to help you navigate the complex—but transformational—opportunities AI is bringing to the world of sales, marketing, and after-sales.

• AI
• Artificial Intelligence
• CRM
• Martech