Skip to content

    Making the Case for DS/ML and Domino in a Recession

    July 15, 2022   5 min read

    Recession? Data science to the rescue!

    A recession looms on the not-so-distant horizon. As a data science leader you may be getting harder questions lately. Where can we trim costs? What’s the value of x, y or z?

    In recent decades I have experienced a few downturns myself. As a customer exec sponsor, avid enabler of sales teams and leader of pricing committees and executive customer advisory boards I have witnessed firsthand the pressure recessions can exert. Fortunately I have also represented several great companies like Domino that help customers be resilient and create significant economic value even during downturns.

    At first glance, data science could be a potential target. Data scientists demand high salaries. They use a lot of data and compute. And many IT and business leaders don’t understand the complete ROI of data science.

    Data leaders invest in downturns

    Digging deeper, though, many leaders actually rely on data science more in a recession. Analytics are essential to identify ways to increase revenue and cost savings. A downturn is also a good opportunity to get ahead in the war for data science talent. It is a smart move to continue to invest and be better positioned for the recovery.

    A few recent articles:

    Increasing productivity to get more value from (and retain) top talent

    I talk to a lot of senior executives such as CDOs and CDAOs. Often compensation is the vast majority of their cost, and they are still fighting for top data science talent in a recession. So keeping top talent and ensuring they’re productive is as essential as ever. Fortunately Domino can be a game changer for productivity, both for data science and IT.  So the Domino customer success team is hard at work  helping customers ensure they get the most out of their investment in Domino. For example, by:

    • Minimizing one-off requests, relying on self-serve infrastructure to automate time-consuming dev ops tasks like configuring servers, moving data around, and managing environments 
    • Eliminating “reinventing the wheel” with automatic tracking and reproducibility of work
    • Saving weeks per model using turnkey model monitoring and not building monitoring pipelines
    • Consolidating multiple tools (Jupyter, RStudio, SAS, MATLAB) onto one platform - no more duplicative silos and support/maintenance burden

    Choosing a workspace in Domino

    One Domino customer was able to repurpose 18 FTEs in IT for more strategic work by simplifying support of data science infrastructure. Investing in top talent doing more interesting work with greater business value is a good recipe for a recession!

    What about infrastructure costs?

    I often hear about how unpredictable and expensive cloud computing bills can be(come). The first antidote to those bills is maintaining a philosophy of openness, flexibility, and customer choice as Domino does. As a kubernetes-native platform, Domino runs on any cloud or on premises. It’s about supporting workloads wherever they are most cost effective. 

    Another fundamental difference for Domino is being the only established MLOps provider that doesn’t charge based on compute usage. In fact, Domino helps reduce your cloud spend, while other vendors are incentivized to drive it up.  Some Domino customers report reducing their infrastructure cost by 40% or more, resulting in $1M+ in savings. How? (I knew you’d ask):

    • Unlike competing solutions, Domino pauses and resumes work easily, so machines don’t run constantly simply for convenience
    • Domino gives warnings and can automatically  shut down long-running workloads
    • Customers use commodity compute with no markup, unlike cloud platforms and Databricks
    • You can set limits on who can use what types of hardware, and how many machines can run concurrently
    • Multiple workloads per machine more efficiently use resources
    • Actionable reports on spend by project, data scientist, or workload type spot the biggest opportunities for cost reduction

    Managing compute spend in Domino

    Amplifying profit through data science

    If you are looking at data science as a cost center you may not be realizing its value. In a recession data science teams can be most valuable by delivering projects that impact revenue and costs… and increase profit. The bottom line is data science can be transformative by:

    • Automating product experiences that increase customer retention and competitive advantage
    • Automating business processes
    • Bringing new innovations to market faster
    • Finding and refining your target customer base 
    • Improving lead scoring and management for better sales efficiency
    • Improving marketing campaigns with better customer profiling
    • Predicting which prospects are more likely to buy or which customers are likely to churn
    • Personalizing marketing content by analyzing past purchases, web browsing and other intent signals

    The opportunities are endless! Which is why companies are investing in maturing their data science for greater value and impact as a recession looms. Fortunately Domino shines in helping customers move beyond experimental projects to more of a holistic system that scales the speed, quality and impact of data science.

    Proven savings and impact when customers need it most

    Of course, there are lots of options out there and you should consult with peers (and hopefully Domino customers!) about how they scaled data science.  Forrester interviewed six Domino customers in a Total Economic Impact and found that on average an investment in Domino realized a ROI of 542% with payback in less than 6 months and more than $22M in benefits from higher productivity, reduced infrastructure cost and increased profits. Here’s where they found the savings:

    • Reduced time to configure compute resources ($9.7M)
    • Improved collaboration in a shared tools environment (>$5.1M)
    • Increased profits through higher model velocity and deployment success rate ($5.1M) 
    • Faster model validation ($1.4M)
    • Accelerated time to onboard new team members ($984K)
    • More efficient model maintenance (>$343K)
    • Retired legacy servers ($141k) 

    The recession means hard choices as businesses position themselves to succeed beyond the recovery. Fortunately for Domino and its customers data science is an area of investment that can drive growth, efficiency and success in the near term and well into the future. Many years from now hopefully you’ll look back at this recession and how it further separated your organization from “the pack” based on wise investments in data science talent, maturity and scale.

    The Total Economic Impact of the  Domino Enterprise MLOps Platform  Download the full study for detailed analysis, customer perspectives, and more. Get the Study

    Rob Smoot

    Rob Smoot is an investor in dozens of early to mid stage enterprise software startups focused on data and cloud. For over 20 years he has built successful teams and businesses as a senior marketing executive and consultant for leading cloud infrastructure, data and management platforms. As VP Marketing at VMWare and Snowflake he grew 3 businesses from early stages to over $1bn in annual revenue. Having been responsible for all aspects of marketing and go-to-market, he has written and spoken in industry forums in many areas of domain and functional expertise. Rob holds an M.B.A. from the Wharton School of the University of Pennsylvania and a B.A. from Brigham Young University.

    Other posts you might be interested in