They say knowledge is power and when it comes to customer relationship management (CRM), certainly that appears to be true. The more information a CRM system has and the more joined up its data is with other functions of the business, the better the outcome.
But CRM systems are living, breathing things that are only as useful as the quality of the data that goes into them. They thrive on fresh information, up-to-date contacts and good quality customer interaction – and they’re certainly not the kind of thing that can be set up and walked away from.
“From an SME perspective CRM, in essence, is keeping your customer data in one place and having it available to all of your employees. And I think what’s very different now compared to what people used to think of CRM years ago, is that CRM data now should drive how companies make their decisions,” said Deirdre Purcell, country sales leader for Salesforce.
“We should all be making data-driven decisions. I think also certainly what we’ve learned over the last year is that every single organisation now has to have a digital strategy, and CRM is [vital] to that. It’s imperative that companies have a digital strategy in order to survive, thrive and continue to grow, particularly after the last 12 to 18 months.”
For Purcell, the role of CRM is to gather customer data together in one place and to connect the company with its customers through that data. If this is done, the result is that the company can make data-driven decisions, and that starts with having a joined-up approach to customer interactions.
“Salesforce allows you to bring in every single channel onto the same platform, bringing all your information into one place and having that create a single source of truth. So our Customer 360 platform, that’s the Salesforce vision of how we help customers on a digital transformation journey,” she said.
“It’s bringing everything from your sales, customer service, marketing, all of your IT departments and your commerce department together – whatever your organisation might have – onto one platform using all the channels a company might connect to customers with, and being able to gather all of that information into one place.”
The thinking here is that while CRM has traditionally been seen as a sales resource. In reality it doesn’t matter if a company employee works in sales or customer service, ultimately they talk to the same customer, and it’s the customer’s experience that matters. As far as that customer is concerned, all employees are equal representatives of the company and it’s their overall experience that matters.
“The Salesforce platform is very much about clicks – there’s no coding in the background unless you’re getting into very complex elements of larger-scale bigger enterprise solutions. So there is no integration: it is one platform. All your information is available to anybody in your organisation you decide needs to have access to it. Obviously there are rules around who sees what and you can set the security rules, but all of that information is in one place,” said Purcell.
“Trust is our number one value at Salesforce. It’s something we absolutely stand by and we have full transparency with our customers in terms of them being able to see all of the system uptime. We have a website called trust.salesforce.com and it is it is paramount to us. Because Salesforce was born in the cloud, we have security built into everything we do.”
For Gary Cullen, sales and marketing director of Provident CRM, it’s important to remember that in Ireland lots of companies are at different stages of their development when it comes to CRM.
“Some companies are advanced in their thinking, and others are still struggling with using spreadsheets and need help to get the process started. There are also companies out there that have deployed CRM in the past but have let them go stale and haven’t maintained them or kept them up to date, which is the cardinal sin with these systems,” he said.
“They’re living applications and need to be kept current. Businesses evolve and change quite rapidly and customer data changes so it’s important that the CRM system does too. You need a strategy for your CRM. Just buying one and switching it on isn’t a strategy.”
According to Cullen, many people think the process of selecting and implementing a first CRM is onerous, so they get project fatigue. Going live seems to be a huge achievement when actually it’s just the start of things.
“But having the system is only the first step in a long journey and that has to be remembered. Having said that, some companies have put in CRM in the past, have structure on their data, have buy-in from their team and really understand the value of what it can do,” he said.
“And now, post pandemic, they’re looking to refresh the technology and take it to the next level. They want to start to use the more modern tools that are now available, such as AI and machine learning, and even just some of the newer integrations that are out there.”
Technology evolves quite rapidly and there are emerging channels for customer interaction that didn’t exist two or three years ago. It’s important that businesses follow their customers, and if there is demand for some new communication channel to be used, it should be used.
“You need to have the newer tools in order to stay current and relevant in the way that you can update and record customer status,” said Cullen.
Obviously, good technology costs and it can be hard for smaller companies to justify spending speculatively, however it’s important to see the bigger picture.
“There are a lot of products out there and, as with anything, they vary from entry level all the way up to extremely advanced. If you’ve never had one before, there are a lot of considerations when going to the market. You should be budgeting €60 to €100 per user per month and that would get you a very good system,” said Cullen.
“You can go cheaper than that but it’s about deciding what’s your three-year plan. These types of technology shouldn’t be just installed and walked away from. You need to be asking if the technology you’re buying today will still be delivering for you in three or four years’ time. So sometimes the year one cost might seem a little high but you have to look at it spread over a three to five-year period and weigh it up that way.”
It’s also the case that some systems are modular and can be scaled up and down based on changing needs.
“So you can take the piece you need today, knowing there is a platform and you can extend into other areas of the platform as your use of a system matures. You can start with a sales element and extend into marketing or into service, you can bolt on machine learning or AI, you can add advanced reporting etc,” said Cullen.
“A lot of the senior vendors have really good offerings now, so there is a lot out there for people to take a look at. You don’t have to spend an enormous amount upfront to find out what these systems can do for you.”
According to Peter Rose, chief technology officer for TEKenable, CRM has evolved in recent years to become something new, which he describes as customer engagement management (CEM).
“This is primarily characterised as the ability to have a conversation with the customer through any medium they want. It could be on the phone of course, or via email but it could also be through Facebook Messenger or all three. An engagement might start on your website and change form multiple times, jumping around,” he said.
“But the system needs to be aware of all those interactions, along with any issues or problems the customer is having.”
Such a system allows the company to stay on top of interactions as they transfer seamlessly from one platform to another without, for example, forgetting a particular insurance quote you gave a customer.
“You need the system to keep track of all your interactions and not lose track of the individual needs of the customer just because the conversation has moved from the web to WhatsApp,” said Rose.
“Where you get AI feeding into this is when you start to drive a lot of personalisation. This is something that has been touted for a long time, but really it’s only coming into its own now. You generate a persona based on the behaviour of someone, where they started their interaction, what parts of a site they looked at, where they moved around to and when they contacted you etc.”
By taking account of all this, and analysing the customer’s behaviour, it’s possible to predict what phase of buying they’re in. Are they kicking the tires, or are they ready to commit? Do they come back to look at the same product three times over a period of time? That kind of help can be invaluable to salespeople and can help make sure each customer gets the interaction they’re looking for.
“Whereas in the past, this was fairly rudimentary, modern customer engagement management powered by AI can also take into account past behaviours of previous visitors and, knowing the outcome of those, statistically compare the current behaviour of visitors to see what the likely outcome is going to be,” said Rose.
“It’s different because you’re using AI to analyse and compare behaviours across populations, as opposed to trying to use defined rules that if a person does X followed by Y followed by Z then they are such and such a type of persona.”
While this kind of artificial intelligence can be enormously helpful, Rose cautions against seeing it as a silver bullet.
“Predictive analytics in AI is just statistics at the end of the day. The classic examples are when it’s used to predict fraudulent transactions, or to ask the question ‘is this customer likely to buy?’. But to do it well you really need a good data set,” he said.
“To do predictive analytics, you need two things. You need a representative dataset to train the machine learning model and you need, and this is really important, you need a relationship between the thing being predicted and the training data.”
In other words, there’s no point trying to use 20 years’ worth of e-commerce data to try to predict the weather if there’s no useful relationship between one and the other.
“Provided there is a relationship, then AI is brilliant because it will constantly tune itself as time goes on and more information becomes available. You can retrain the model on that data, and you get more and more accurate in terms of what is being predicted,” said Rose.
“One other thing that’s really interesting is that the tools needed to make use of machine learning and AI have become a lot easier to use. At the start with these tools, you needed a PhD to configure them in order to get really powerful results.”
That’s no longer the case, and according to Rose, it’s now possible to make use of these tools much more easily.
“The type of tools we use now include AI packaged as a consumable service. This is stuff you can build on a website. You can go to a website like Microsoft AI builder, for example and then go to your site and it’ll ask you what you want to do. Do you want to predict multiple outcomes, single outcomes? It doesn’t bother you with how it works. It just wants to know where to find the data it needs and what you want the outcome to be,” he said.
“And it asks you in normal business English. That makes it much more accessible and easy to embed in your systems. And that’s really key: AI is becoming much more accessible.”