Enterprise hits and misses - ChatGPT Enterprise launches, Google Cloud Next wraps, and composable ERP gets real

Lead story - Is composable ERP ready for prime time? Customers weigh in

How did AI get bumped off the headline spot? Because Phil wrote one of the most compelling ERP posts of the year: MACH early adopters beat a path to the composable future of ERP.

The biggest strength of this post? Instead of another futuristic musing about composable ERP, Phil shares real world use cases from the Mach TWO conference. The use cases aren't the same, either:

1. Build your own - One customer (Boohoo Group) built their own composable ERP system.
2. Tie to customer-facing systems - Nudie Jeans "uses a Sitoo point-of-sale system and the Centra headless e-commerce platform as its customer-facing transactional systems."
3. Adapt your current system into services - Emma Sleep "Emma Sleep has worked with its ERP vendor to adapt the system so that it provides specific services within the overall composable architecture.

I don't think most enterprises will choose to build their own composable ERP system, but it does put ERP vendors on notice: watch out, or your transactional footprint will shrink,  as customers build out composable apps, microservices, and pull in cloud services. What do all these customers have in common? A push towards modernization that does not center around their traditional ERP vendor. As Phil puts it:

Rather than dividing the IT landscape into back-end monoliths and front-end platforms, why not build a single architecture composed of autonomous services? Within this unified landscape, each transactional service forms part of the system of record.

I agree 100% with Phil's critique of the so-called bi-modal approach to IT.  This piece illustrates how customers can modernize at their own pace, without ERP rip-and-replace on the one hand, or all-in-for-the-composable-enterprise fantasies on the other. However, I have one nit to pick. In the intro, Phil writes:

Some early adopters are starting to replace conventional ERP systems with more composable alternatives.

I think I understand why Phil chose the world "conventional," but I would have chosen the word "legacy." Why? Because a major reason why ERP systems are criticized in this piece is not because they are conventional (monolithic), but because of dependence on batch processing in a real-time world. I can think of a number of ERP vendors that would cry foul that they are still dependent on batch processing (a customer might still choose to do it, but the technical barrier isn't necessarily there, and options like 'micro-batch' are becoming more commonplace). Other ERP vendors go so far as to claim they are indeed composable, though as per Phil's prior pieces, those claims likely wouldn't pass the full MACH inspection.

As Phil alludes to, some prominent ERP vendors are rolling out cloud services models that allow their older customers to upgrade to cloud services in place, without having to do a major ERP upgrade. That's not radically different from some of the use cases described here.

However: I am rooting for these MACH-oriented composable ERP use cases. These stories show customers: the art of the ERP possible isn't a slide deck; it's real. In turn, this puts pressure on ERP vendors to make modernization easier, bolster APIs across releases, and accelerate their own path to composability.

diginomica picks - my top stories on diginomica this week

• An important development - US regulators revisit copyright for AI - George files an important update. This unfolding story isn't over - not even close: "The current legal landscape around generative AI content is still a Wild West, which might be considered an opportunity or theft, depending on which side of the equation you are on. Some firms are quickly rushing in to hoover up data at epic scale to train better models."
• Can enterprise tech redeem itself with generative AI? Vishal Sikka on doing AI right, and avoiding generative AI snake oil vendors - In the second part of my interview with Dr. Vishal Sikka, founder and CEO of Vianai Systems, we discuss a novel question: can enterprise tech redeem itself with generative AI? Sikka also shares the top questions customers should ask - to separate solid generative AI vendors from pretenders.
• How Ahold Delhaize is using Wi-Fi roaming technology as a platform for data-led innovation - as Mark Samuels documents, don't fear the in-store + online customer: "The margins can be tighter in retail than other sectors. But that also forces you at the same time to really look carefully at cutting-edge technology." Sidenote: interesting privacy/opt-out/in-store personalization adventures ahead - for all of us.

Vendor analysis, diginomica style. Here's my three top choices from our vendor coverage:

• Boomeranging back - a "modern version" of Salesforce delivers expectation-beating profit margins - On the almost-eve of Dreamforce, Stuart assesses the earnings (and "boomerang" leadership (re)hires. In addition to the multi-cloud boost, this jumped out, re: pressing data needs: "Mulesoft momentum was cited by CFO Amy Weaver as a primary driver of revenue growth." Also see, by Chris: There's never been a better time for consultants! Examining IBM's new AI alliance with Salesforce.
• How Swiss Re goes about de-risking insurance company risk with TIBCO - Martin details a fresh TIBCO use case: "More than a third of Swiss Re’s 15,000 or so employees are now regularly using the connected data assets that have so far been created as part of the strategy, and that process will continue while there is data still to be collected and analysed."

Google Cloud Next '23 coverage - the show is in the books, and Derek has the frequent flier miles to prove it. Some coverage highlights:

• Google Cloud Next ‘23 - CEO Thomas Kurian on AI pricing, sustainable models and the competitive landscape
• Google Next ‘23 - Estée Lauder Online EVP says generative AI ‘FOMO’ is a trap

Via Derek's Kurian sit down:

Google Cloud wants to provide the building blocks for companies to leverage AI according to their needs, via its Vertex platform. Vertex provides not just a collection of models, but a number of foundational services to make AI effective and trustworthy - such as information retrieval, search, conversations, statementment, grounding, watermarking, synthetic data generation.

I find the AI pricing conversation/debate fascinating - as Derek reports, Google currently plans to charge $30 per user per month for its Duet AI assistant services. AI pricing is a topic I'll return to throughout the fall, with rigor.

A couple more vendor picks, without the quotables:

• Samsara revenues soar as cash flow milestone is reached - Stuart
• Citizens Advice tackles cost of living crisis with Freshworks, contemplates AI potential - Madeline

Jon's grab bag - Mark Chillingworth delves into a localized use case in Runnymede - from Magna Carta to simplified public services. Barb explores AI's impact on B2B sales in 6sense study demonstrates AI’s potential for B2B revenue teams. Chris looks at why NASA applies AI with an industry/private approach in How NASA is using AI to search its own data universe. Finally, Stuart assesses the state of generative AI for the enterprise in Salesforce research finds business buyer and consumer trust levels in AI decline despite the generative hype cycle: 

80% state it’s important to have people validating the output of the [AI] tech. That human validation aspect is cited by 52% of respondents as a factor that can increase customer trust in AI, along with more customer control (49%), third-party ethics reviews (39%) and additional government oversight (36%). But the main thing that would make a difference is greater visibility into how AI is being used by an organization, ranked number one by 57% of respondents.

Best of the enterprise web

My top seven

• Everyone wants responsible AI, but few people are doing anything about it - Joe McKendrick cites an AI data discrepancy: "73% of business leaders say AI guidelines are indispensable. However, just 6% have established such guidelines in their companies." McKendrick cites useful AI project tips from Google, including "design your model using concrete goals for fairness and inclusion," and "stress test the system on difficult cases."
• As ChatGPT goes Enterprise, here are Ten GenAI Reality Checks you need to take - HfS Research issues a gut check as ChatGPT Enterprise is launched. However, I'm not convinced ChatGPT will win the enterprise market despite its impressive consumer adoption. Why? 1. I'm not convinced ChatGPT's base model will excel at enterprise problems, despite changes in "context windows" and customization options. 2. I'm more inclined towards the industry LLM play, which Constellation's Larry Dignan lays out in Get ready for a parade of domain specific LLMs.
• Nvidia On the Mountaintop - Ben Thompson makes sense of NVIDIA's AI-driven surge, and rethinks his past positions on the company.
• 4 Key Observability Best Practices - I had to turn over a few rocks to find good non-AI posts this week, but I liked this one from The New Stack.
• 5 ways CISOs can prepare for generative AI’s security challenges and opportunities - Louis Columbus looks at the emergent application of generative AI to IT security. Can generative AI defense keep up with generative AI-based security attacks? We'll see.
• No, Virginia, AGI is not imminent - Gary Marcus calls BS on Artificial General Intelligence hyperbole.

Whiffs

Who knew that scientologists don't want you fixing your own gadgets?

You have to be really drunk on the generative AI Koolaid to think that mushroom classification is neato use case:

I took another swipe at disappointing enterprise event thinking:

But I'll pay my penance with my next article, which will lay out some practical ideas for the better. See you then...

If you find an #ensw piece that qualifies for hits and misses - in a good or bad way - let me know in the comments as Clive (almost) always does. Most Enterprise hits and misses articles are selected from my curated @jonerpnewsfeed.