Given that the AI market is expected to be worth $390 billion by 2025, it’s a safe bet that most partners will be talking to customers who have AI-fever. Of course, each business will be at a different stage of adoption. Some will have only just started to dabble in AI-enabled services, like chatbots or Robotic Process Automation. But many midsized or large enterprise customers are already using AI as an engine to power transformation, leveraging advanced analytics and machine learning techniques to optimise their operations and refresh their business model.
For those customers, the way they’re using data is fundamentally changing and therefore so too must the way they store and manage it. And for channel partners, that means the way they talk to customers about data storage is going to have to change in step.
A 2019 report found that 40% of organisations that had made a significant investment in AI technology hadn’t experienced any real business gains from it. So even if customers aren’t proactively approaching their partners with AI-related woes, there is an opportunity for partners to identify opportunities to educate their customers and deliver added value by understanding their frustrations and suggesting technological solutions that take these into account.
But most account managers understandably aren’t AI experts. While partners don’t need sales teams made up of individuals with doctorates in data science, being well-versed in exactly how AI workloads will change a customer’s data storage needs will set partners apart from the crowd at a time when proving value and building relationships is paramount.
So how exactly is AI changing the data storage game, what are some of the bumps in the road customers will likely encounter and where can partners help?
AI workloads require next-level infrastructure performance
To train their self-driving cars, Tesla drew on over 1.3 billion miles of driving data. Developing algorithms for ambitious AI projects requires the processing of huge data sets. Securing all that data in a way that is effective for training algorithms and running advanced analytics requires high-performance architecture that is scalable, reliable and enables data to be easily discoverable – especially unstructured data, which is what is typically analysed today and isn’t easily searchable.
There are several factors partners should bear in mind when considering storage solutions for customers with specific AI requirements. Machine learning (ML) or AI training techniques require greater infrastructure efficiency. A storage system will need to read and reread data sets regularly, often at random, which rules out tape storage options. Latency is also a critical factor. Because data is being read over and over again, a system with low-latency can cut training time down dramatically, ultimately driving down costs and increasing efficiency. Likewise, when processing such huge volumes of data, throughput is critical and some systems will struggle to keep up.
As a general rule-of-thumb it’s said that data scientists spend around 80% of their time just preparing data. So when selling a solution, partners should be able to convey to customers just how the right storage strategy will maximise the productivity and effectiveness of their data teams, having a direct impact on the success of projects.
But what the “right” solution will look like depends on the customer – their budget, security needs and desired goals. And that’s where channel partners can really prove their value.
The opportunity for channel partners
For years now, channel organisations have been cottoning on to the fact that they must shift from a vendor-led approach to a customer-led one if they hope to drive new revenue opportunities. Given that, according to one study, almost half of buyers have already identified specific solutions before they’ve even spoken to a sales rep and only about 1 in 4 consider sellers to be a top resource for solving business problems, partners can no longer just sell products from vendors. Most now act as consultants, putting customer experience at the heart of everything they do.
A blanket, one-size-fits-all approach to distribution no longer cuts the mustard. Partners must stay close to their customers, nurture relationships, learn about their business and provide highly-tailored solutions that will deliver business results.
Determining whether the optimum storage architecture is on-premises, in the public cloud or a combination of both will depend on multiple factors, including the type of projects the customer will be running – both now and a few years down the road. For larger enterprises, long-term, large capacity projects will likely be developed on-premises while public cloud is better suited to smaller-scale, less demanding work. For some midsized customers stepping-up their AI projects with the potential to scale-up in the future, flexibility will be crucial. Vendors are now starting to offer flexible consumption models, enabling greater cost transparency, lowering overheads and a creating a pay-as-you-go approach that is ultimately a lot lower-risk for customers that might have a tighter budget or a less concrete AI strategy.
With such an extensive menu of options available, partners are an indispensable resource for customers overwhelmed by choice. Defining and implementing a storage strategy that is aligned to their AI goals and will enable them to get the most out of projects requires up-to-the-minute market knowledge that only channel partners can provide. So, if customers aren’t already coming to their partners to ask about how their AI projects are going to impact their data storage requirements, now is the moment for partners to connect the dots for them and drive those conversations.
Join the Hitachi Vantara Partners group on LinkedIn to keep up with our latest partner news and discussions.