Hi everyone,
It’s that time of the quarter - Earnings!
And, the most notable earnings release on our radar was Nvidia. Now, one of the 5 most valuable companies in the world - beating out both Amazon and Google!
With a $2 Trillion market cap - Nvidia’s earnings release on Thursday was nothing short of a blockbuster, especially considering the company was worth less than $1 Trillion a year ago.
Nvidia earned over $60B in revenue in 2023, up from $27B in 2022 - a 125% YoY growth. More importantly, their net income increased to $30B, more than their 2022 revenue. Up from $4.4B net income in 2022 - a 580% increase in net income!
As many of you know, Nvidia is a chip manufacturer, and historically, would manufacture the GPUs, graphics processing units. The GPUs are very important for playing video games and that’s been a big part of Nvidia’s business for many decades. However, it turns out that the processing needs from AI supercomputers also need GPUs.
Nvidia’s cloud business unit had the largest growth and was a major contributor to its 2023 performance. Similar to AWS, Nvidia’s cloud business offers AI processing power which leverages its Nvidia GPUs as-a-service.
The five industry’s highlighted for being power users of Nvidia’s cloud services were:
Automotive
Healthcare
Financial Services
Consumer tech monopolies like Amazon and Google
Enterprise tech companies like Cisco
So, who loses in the AI revolution?
In industries like healthcare and financial services - two of the big 3 verticals for Nvidia’s cloud business, there are a myriad of white collar jobs, many of these tasks are outsourced to other countries. And, many companies such as BPO, business process outsourcing, have hundreds of thousands of employees in countries like India, the Philippines and other English-speaking countries that have much cheaper labor. The jobs and tasks that these people are doing aren’t terribly complex, but the tasks have historically been too bespoke for a computer to automate them - until now.
Take TaskUs ($TASK) as an example, they went public in 2021 at roughly $30/share and are now at $12/share. Key examples of outsourced services would be customer service, reviewing “harmful” content on social media, fraud and identity verification for banks, etc.
Revenue growth has flat-lined for TaskUs; but the company is still maintaining its margins and net income. It’s probably using AI to automate a lot of the tasks they have used humans for!
Eventually, AI will be used to automate more and more of these kinds of services. Will TaskUs be able to pivot the business? Maybe - but it’s going to be difficult.
AI in B2B Distribution: The Three Pillars and Two Buckets
AI should obviously help distributors run a more efficient business; however, AI needs data, connectivity and workflow to effect change.
AI needs data, connectivity and workflow to effect change.
If AI can’t access a distributor’s data - it can’t do its job. Therefore, AI needs connectivity to access data inside a distributor’s ERP, WMS, TMS and other various systems. Once AI has connectivity and access to data - it needs to interface with a distributor’s employees via workflow tools. If the AI can’t help the purchasing manager make better decisions, then the benefit can’t be captured by the distributor.
Certain tasks will be automated almost entirely - but a lot of efficiencies will make existing staff more productive. We call these hard-cost savings versus variable-cost savings.
The former is a turn-key, point solution. A distributor plugs it in, the AI tool performs its tasks and savings are realized. These costs are extremely objective and measurable - one example is Loop, the AI company Applico Capital invested in last year to help distributors automate invoice audit.
The latter bucket are variable-cost saving AI solutions. These solutions make employees more productive and help them make better decisions in less time. These costs are a little harder to objectively measure because they vary by organization to organization. Depending on how the distributor’s staff operates or what verticals they focus on, the cost-savings and efficiencies are more variable and the use-cases vary. One example is Tyson Foods. They utilized computer vision to increase inventory tracking accuracy by 10%. This accuracy translates into optimized inventory levels, mitigating the risk of stock-outs or overstock situations - and reducing the efforts needed by warehouse staff to properly monitor inventory levels. How much exact savings are realized by a 10% increase in inventory tracking accuracy? It’s hard to say exactly, but one can imagine that Tyson has pretty high confidence that the SaaS fee they are paying to the AI company is well worth the savings & efficiency gains from increased inventory accuracy.
The Decentralized Distributor and AI
Distributors are generally very decentralized - with branch and regional P&l’s. Over the past couple decades, we’ve seen centralized IT functions help to automate more and more processes; however, the benefit from centralization is harder to realize within distributors that have high, inorganic growth from M&A. That’s why we believe the value of AI in distribution will be twofold: finding efficiencies at the branch level and helping centralized IT departments with the three pillars: connectivity, workflow and data.
Keep reading with a 7-day free trial
Subscribe to B2B Distribution Technology to keep reading this post and get 7 days of free access to the full post archives.