Is your CRM strategy advanced enough for artificial intelligence?

In his latest novel, Machines Like Me, Ian McEwan reimagines what 1980s Britain might have looked like if certain iconic events had swung a different way: Argentina winning the Falklands war; Margaret Thatcher battling Tony Benn for power.

And then there’s the catalyst for the book – what might have happened if Alan Turing had achieved a major breakthrough in the field of artificial intelligence.

As with most visions of AI’s future – or hypothetical past – the author speculates a dystopia in which sentient, anthropomorphous robots are embedded into our everyday lives.

The pros and cons of AI are debated at an ethical and moral level; with less credence given to the practicality or the technological constraints.

Reality check

Whilst undoubtedly a necessity for a science fiction novel, in the real world, this lack of practical consideration for AI and the related constraints has been responsible for a huge amount of hyperbole in the enterprise technology market; especially in relation to CRM. If you believe in some of the vendor doctrine, AI has magical, transformative qualities that must be plugged into without haste.

So what’s the reality?

“For me, where we’re at with artificial intelligence is similar to where we were with social CRM, seven or so years ago,” says John Cheney, CEO at Workbooks.              

“The reality for most organisations is that AI is not something that they’re necessarily geared up for or actually need. We’re in danger of pitching AI as being the answer to pretty much everything. For most of our customer base, AI is some way off. Maybe in the future they’ll be in a position to utilise AI, but based on where many of them are at today, it shouldn’t be on their radar.”

This is a refreshingly candid outlook on the state of a market that is seemingly transfixed on artificial intelligence. A 2017 Statista study estimated that between 2017 and 2021 an additional 394 billion US dollars would be gained in the United States through the adoption of AI in CRM.

Many organisations heeded this call. In IDC’s 2017 AI/CRM Economic Impact survey, 28% of respondents said their organisations had already started using AI. An additional 41% said they planned to adopt AI in the next two years.

What some converts failed to do during that period was ask whether they were really in the right position to plug into AI to start with.

“If you focus on the mid-market – which is where we operate – and even at large enterprise level, many companies just don’t often have their customer information in one place,” says Cheney.

“Finance data may be on one platform, operational data on another…etc. Those disparate data sets make AI virtually impossible, as you need that data to be working together to do the kind of analysis AI is there to do.

“In those instances, AI has the capability of offering value in pockets of the CRM journey, no doubt; most notably in the marketing department, because we can fairly effectively analyse customer interactions, email opens, web journeys, etc, and use that data to form some analysis on typical customer journeys and workflows. That’s ultimately what AI does.

“But if you think about things like lead scoring in sales, that’s the start of using logic to apply rules to say if A happens then the likelihood is B will happen as a result… but that’s reliant on transactional data on your platform alongside typical sales data and if you don’t have that, you can’t lean on the necessary AI for help.”

Ducks in a row

The hype around AI can often mean many brands are striving to benefit from AI without getting the fundamentals right first. A 2018 Forbes Insight study surveyed over 400 marketing leaders to find that, even in a department that’s seen as being the greatest benefactors of AI, only 13% were confident of being able to use all of their available customer data effectively.    

Customer Data Platform Institute research also highlights that many marketers see a lack of ‘single customer view’ as directly inhibiting their ability to perform effectively.

In those instances where unified data is holding performance back, it’s understandable that some businesses might turn to AI as the answer. However this mode of thinking, Cheney adds, is compounding the problem.

“Before looking at AI’s role in CRM, you have to ask what the objective is of CRM itself, in your business. We always talk about four key outcomes – the first being revenue growth: how you can sell more to your existing customers and drive new business; the second being reducing operational costs: streamlining how you do things within your business; third is about improving customer experience; lastly it’s about having better information to make more informed business decisions.

“AI can potentially offer help in all of those areas, but the reality is in order for AI to be effective, it needs huge data sets to analyse and for most organisations, their CRM platforms just don’t contain enough data.”

Cheney believes the most practical approach to an AI-led future in a business is to first focus on how to align and unify all customer data; no easy feat for many.

“Businesses need to consolidate their platforms into a smaller number of systems so their data can flow freely through the organisation. Having any kind of customer view is nearly impossible without this.”

In many cases this can lead to achieving many of the objectives AI is often said to be tasked for.   

“You can get some real insights from just having the data in one place – at Workbooks we are still bringing customers onto our platform who don’t know which customers are buying which products from them. You don’t need AI to solve that issue. That’s job number one. Get your CRM in place, get it understanding your customer journey, and then you can start thinking about AI.”

Humans still pivotal

According to Sales Hacker, the average employee spends four hours per week entering data into their CRM, equating to 25 days per year.

Yet since Gartner’s 2001 study found a 50% CRM failure rate, study after study has reported failure rates between 30% and 70%. A primary cause of failure still remains the lack of use among personnel, especially outside of the sales team.

As commentator Jenna Dobkin wrote for MyCustomer last year, “people would rather spend their time working around their CRM, continuing to manage their contact relationships via email and retaining customer information on spreadsheets, rather than change their work habits”.

It’s no coincidence that the greater the discussion becomes around artificial intelligence, the greater we highlight the necessity of human input in order to make CRM function correctly.

“Businesses need to bring people together on the data discussion. In larger enterprises, you have whole departments that might not be in the same building and there may not be a culture of speaking to one another and sharing data,” adds Cheney.  

“There’s a real requirement to proactively look across the business and establish how you share information, and work together more effectively.  Those data silos do exist in smaller businesses too; it’s about building a culture that has CRM project team that can represent the different needs of different parts of a business but also identify where silos exist that would benefit from better information sharing.”  

In Machines Like Us Ian McEwen asks the existential question of why humans need AI at all, when so much of what we’re so often trying to achieve is merely to match human capability. As Cheney adds, it’s a concept that can be applied to rational thinking around the near and future use of artificial intelligence in CRM.   

“You can get a long way without AI – a quantum leap – there is a trend in the tech sector that we’re very good at taking a concept and try to productise and sell it before thinking deeply about whether it’s a true form of that concept.

“We are far too quick to market with things that probably won’t add value to many prospective customers.

“Ultimately, my advice is that in the short-term – AI will offer most organisations very little value, it’s only in the long-term and only after a business has made some significant changes first that we should consider its true value.”