Carolyn Parent is CEO of Conveyer, an AI data transformation platform that extracts value from underutilized, proprietary unstructured resources, transforming them into high-quality data that can be used to power AI projects. A tech veteran with 25+ years of CEO- and board-level leadership experience, Parent recently served as an Entrepreneur in Residence at HearstLab, and has raised over $50 million for four different companies.
Data migration hurdles hold back AI innovation and hurt both providers and customers.
The CRM industry is booming. According to Forrester, 57% of companies will boost their CRM investments over the next year, with 75% of leaders seeking customer experience improvements and hoping for big efficiency gains.
Part of the reason is that the CRM space, like many others, is being transformed by new AI technologies. One study projects that GenAI alone will help service agents to cut response times by 30%, and accelerate marketing campaign creation by 35%. According to Forrester, in fact, CRM and CX are among the key areas where AI will start to drive real impact in 2024.
To unlock the power of AI, however, organizations need to leverage customer data efficiently. That’s where the problems start. While CRM platforms are great at ingesting new data, they almost uniformly do a terrible job of activating legacy data recorded using other CRM platforms. That leaves organizations struggling to use much of their most valuable data, and creates customer lock-in that ultimately holds back growth for everyone.
CRM depends on ugly data
Much like healthcare records, a huge amount of CRM data exists in ugly, hard-to-leverage formats. A big part of what makes CRM systems so powerful, in fact, is their ability to onboard that data and make it actionable for business users. Support tickets, customers call transcripts, technicians’ calls, service data — together, they constitute a treasure trove that organizations need to leverage to drive insights, win sales, and deliver better support.
The problem is that this “ugly” data usually winds up formatted in ways that only make sense for the specific CRM platform they’re first captured in. Try to migrate legacy data across to a new platform, and you’re usually out of luck — and the richer and deeper your legacy datasets, the harder it is to process or port them into new CRM systems in any meaningful way.
Even CRM giants like Salesforce struggle to unlock this treasure trove. At best, the major CRM players let customers dump their legacy records into their new CRM system as a single block of undifferentiated and unstructured data. No real effort is made to organize or process that data, let alone integrate it into the platform’s existing data structure.
That’s the equivalent of dumping data into a bucket, but never being able to look inside. It might technically live in the new CRM system, but there’s no way to activate it to drive value — or to use it to enrich the new AI tools that companies are increasingly counting on to deliver efficiencies and value for customers.
The trouble with neglecting legacy data
Why does legacy data get neglected? It’s partly that big CRM players are caught between conflicting priorities. They want to make it seamless for companies to switch to their platforms — but they also want to keep existing customers locked into their platforms. If perceived friction, or fears about losing easy access to data, keep customers from shopping around, that’s a clear win.
For CRM customers, however, letting legacy customer data fall by the wayside can be a big problem. First, legacy data creates compliance risks, because it’s still subject to data privacy rules. If a customer wants their data deleted, for instance, you need to be able to process that request quickly and reliably. That’s far harder to do, of course, if the relevant data is dumped in an inaccessible bucket that your current CRM system can’t access, monitor, or operate on.
Secondly, unprocessed legacy data can’t be used to power AI innovation. While both current and legacy data might contain insights about customers’ common service issues, for instance, it’s hard to unify them in ways that allow those insights to reinforce and add value to one another. The result: years or decades of valuable historical customer data goes untapped.
As companies begin to adopt GenAI-based CRM tools, meanwhile, data migration will grow even more challenging. If you’ve fine-tuned your company’s customer service AI models on one set of data, what happens if you migrate to a new system and can no longer easily retrain your AI models? Losing access to legacy data could degrade emerging AI models and de-incentivize migration from one CRM platform to another.
Ensuring data interoperability is tough even when organizations are motivated to agree a common data standard. In competitive industries like CRM, big companies want to defend their data advantage. That means there’s little prospect of everyone getting together and agreeing new protocols to let customers easily port data from one platform to another — or to let them bring their historical customer data into the mix.
How to put things right
The solution is to cut through the Gordian knot. Instead of using data to dig moats around our businesses, we need to help data to flow to where it’s most needed — and compete not on the walled gardens we’ve built, but on the amazing things our platforms let users do with their data.
In an ideal world, perhaps every business would put the perfect CRM solution in place when they first incorporate, and let their data and workflows grow up organically alongside that solution. In practice, of course, organizations’ needs change, and so do the available technologies. Inevitably, that sometimes leads companies to upgrade their CRM providers.
Using new AI technologies, fortunately, it’s now possible to breathe new life into legacy data and automatically unify the competing structures used across current and legacy datasets. Algorithms and natural language processing can restructure and transform historical datasets to make them accessible to new CRM platforms, streamlining migration and radically elevating the value it’s possible to create with new AI models.
For CRM vendors, that might sound like a mixed blessing. It’s true that removing the friction from data migration will create more fluidity in the marketplace, and make it easier for customers to switch vendors. But that increased competition will be good for everyone in the long run. Companies that use data — including legacy data — to drive real end-user benefits will be rewarded in the marketplace, and emerge as leaders as the CRM sector joins the AI revolution.