deal architect

Year-end musings: Of Marketwatchers and Sheep

Third of year-end posts I am running this week.  The other two were about enterprise software and outsourcing trends.

I recently saw The Big Short. Well done, actually funny, adaptation of the book. With so many things in the movie that make your  jaw drop you could easily miss the cameos where characters from Standard & Poor’s, the SEC and The Wall Street Journal act casual or even defend the growing mortgage loan crisis that brought the global economy to its knees.

It reminded me of Chapter 11 of SAP Nation titled Market Watcher Omissions. During my book research, I was stunned to find so little coverage of trillions in spend, much of it wasted. I reached out to analysts, journalists, user groups, academia, regulators and other marketwatchers for my research.

As I wrote

“..I had a gnawing sense of “How was this allowed to go on and on?” The initial IT failure of the Obamacare-related Healthcare.gov got a relentless amount of media, political and business scrutiny. That was reported as a $1 billion project (even after the overruns). In contrast, the SAP economy has had significantly more write-offs and waste. Why has it not seen anywhere near the scrutiny? 

It’s not just around SAP. I have heard in the last few months a couple of tech executives call financial analysts “sheep”. There is little independent scrutiny of results and some vendors are known to feed them “suggested questions” to ask of competitors.

As I continue my research on automation and impact on jobs for my next book, I see so many glib, anecdote based comments from analysts and bloggers. There is so much labor statistics data out there, I wish they would do their homework before they just opine on stuff. Then there are analysts who relentlessly tweet from industry events basically regurgitating what tech execs are presenting on stage. That passes as analysis?

I think we would all do well to emulate Michael Lewis, the author of The Big Short, Moneyball and Flash Boys among others. He does lots of research – for months and years – takes strong positions, bases his stories around real life but previously undecorated folks like Billy Beane and Brad Katsuyama.  His books have tremendous impact, yet I like the fact that he is happy to be be still called a “journalist.”  No high falutin’ terms like analyst for him. And I doubt anyone would dare call him “sheep”.

Year-end musings: The changing outsourcing landscape

Second of year-end posts I am running this week. Yesterday I wrote about enterprise software trends.

To me 2015 was about how three A’s – Amazon, Analog, and Automation – are changing the outsourcing landscape, and looking forward, how these drivers will likely accelerate changes in the coming years

Amazon – With Amazon now more open about economics in its AWS business, every outsourcing executive should accept three realities. Firstly, the old business model cannot survive. Customers now expect more continuous price and performance improvements. Ok may be not 3-4 times a year like AWS has delivered, but certainly not once every 5 years that most outsourcers got used to. Secondly, they will have to rethink capex. Most outsourcers invest heavily in recruiting and  staff training, not in data centers or automation. That game has changed. Finally, they need to expect more “left-field” competitors.  If they only focus on competitors they see at software conferences or NASSCOM events, they will be in for more surprises. Amazon is just the first “outsider” they have seen from what I can tell from my industry conversations. Let’s not forget AWS targeted a narrow sliver of the infrastructure outsourcing market, but it outclassed the incumbents with their outdated data centers and contracting models. Plenty of application, BPO, systems integration areas offer similar opportunities for more efficient outsourcing.

Analog – or its antonym, Digital, has become a buzzword for most outsourcers. Most sophisticated customers I have followed (since writing The Digital Enterprise couple of years ago) find their digital edge comes from making previously analog products, services and business models “smarter” with embedded or enabling technology. Since those projects are so competitively strategic, most of them only selectively use outsourcers, and tend to use boutique design and contract manufacturing shops. So, most outsourcing “digital” experience I find tends to be in the back office or in some cases around social marketing, but not much around digital product engineering or manufacturing. If this does not change shortly, many outsourcer digital practices will essentially be spray painted ERP practices and will struggle to differentiate.

Automation – this is becoming the new buzzword for the industry with analysts like Frank Casale and Phil Fersht talking about “robotic process automation”.  As I continue research for my next book about automation and impact on jobs, I am finding, outsourcers are offering to consult or even reengineer customer processes with an automation angle. But most are talking “machine learning” when automation means wearables, robotics, drones, 3D printing and other emerging technologies for many customers. Few outsourcers have such operational automation depth. And when it comes to machine learning, their skillset is puny compared to what is going on at consumer tech giants like Google and Microsoft. Also, from a credibility point of view, outsourcers need to showcase what they are doing to automate their own labor intensive, especially younger staffed, operations. As an executive told me “I would feel much better about IBM’s Watson if it did not come strapped to a 50,000 person outsourcing business”. It’s not just an IBM issue – it is a much broader industry issue, and a tricky one as it balances pressures from various countries to continue to grow youth employment opportunities.

The three items may sound daunting but the industry has repeatedly shown the ability to morph with the times. Plenty of morphing on the cards in the next couple of years. 

Year-end musings: Speedier Cloud Boats

First of year-end posts I will run over the next few days.

This time of the year, my wife Margaret likes to take the family to the Caribbean, usually on a cruise. The island breezes are gentle, the rainbows at sea are spectacular, the pace is extremely relaxed. So much to like, but you have to accept the slow sea travel pace. Top speed of 20 knots, usually much slower with the resistance from the headwinds and currents.

It is always a reminder to me about the state of enterprise software. That cloud customers – in spite of impressive success of Amazon, Workday, Salesforce, NetSuite, and others – are still islands in a sea of on-premise customers.

This year as I wrote SAP Nation 2.0 and looked at slow migrations in the J.D. Edwards, Oracle, Microsoft and Infor customer bases, it was a striking reminder of what Bill McDermott of SAP has called his and could be said about other on-premise customer bases : they are unassailable like Fort Knox.

The problem is customers paid gold prices for on-premise implementations, but against cloud benchmarks, what is stored in that Fort Knox is merely gold plated and needs to be dramatically depreciated. And with most high-profile security breaches in the last few years in on-premise settings, Fort Knox is too ambitious a term to use for them.

Several things need to happen.

CFOs and audit firms need to become more demanding in accelerating write-offs of hundreds of billions of on-premise investment costs that are still on so many customer balance sheets. The fair market value of most on-premise cost components – software, compute, networks, systems integration – has dropped significantly. Realize it is a very tough decision, but many companies wait too long and then have to take spectacular end of life write offs when they should have accelerated the depreciation over the last few years and could have made the cloud switchover decision easier.

Wall Street needs to start valuing vendor on-premise revenues at public utility comps – predictable, but low growth and under margin pressure from third party maintenance and other competitors. On-premise vendors bitch and moan that cloud vendors are over valued given their low profits. Frankly, it’s the other way round. If Wall Street were to value on-premise revenues at much lower multiples, you will see those vendors get a lot more serious about migrating their customer bases to the cloud. Less than 10% of SAP’s new S/4HANA customers are expected to adopt its public cloud option.  You think SAP would behave differently if Wall Street valued the “private cloud” revenues much lower?

Rather than waiting on audit firms and Wall Street, cloud vendors need to do their part. Their functional footprint has to keep growing, especially when it comes to industry functionality. They have to automate migrations to the cloud – too many are still reliant on the on-premise implementation paradigm and systems integrators from that world.

The good news for cloud vendors is there are hundreds of thousands (easily) of on-premise customers waiting to be converted. They just need speedier boats to get customers there, not the relaxed cruise ship I was just on

The whole world as one “area code”

This time of the year, I am reminded of my Brazilian-American friend, Karin Morton. With family spread around the world she said her father would have palpitations, as she was growing up, with all the international phone calls at $ 5 a minute at holiday season.

Bell Labs thought up the 3 digit US area codes in the 1940s. They have spread around the world in the form of country and city code through ISO and other standards. The codes made long distance calling so much easier but telcos have also used them to terrify us with big bills.

Skype changed the rules, and with Vonage now, I don’t hesitate to call anywhere, anytime as their monthly global plan covers most landline calls around the world. Mobile phones, though, continue get us with roaming calling and data charges.

That’s gradually changing. My new Google Fi plan, when using local Sprint and T-Mobile partner networks, only costs 20c a minute from most of the world. On wi-fi, it costs nothing. Sprint has a plan called Open World where a number of S.American countries are counted as part of your US plans. Their rates for other countries are also improving.

But as with everything, there are always gotcahs. Sprint reminded me “If your travel plans include a cruise and you do not wish to incur these costs, you may place the device in airplane mode until you arrive at your final destination.”

70 years later, telcos still love playing Grinch.  The world keeps getting smaller and the dream of every call as a local call continues. Just may take another 70 years.

Unit4 “self-driving” UX evolves

I was excited when Unit4 shared this summer early details of its machine-learning driven thinking (shades of Microsoft Cortana applied to enterprise scenarios) – I titled it “AI is the new UI”

Yesterday, they shared couple of use cases from their services vertical – one around proposals (an area which take a lot of effort in that sector, yet is repetitive in so many elements) and project risk analysis (certainly an area which has stubbornly plagued IT projects which seem to be almost guaranteed to miss time/budget estimates). Early implementations are proceeding at customers like Groupe ALTEN, which focuses on technology projects, and Parsons Brinckerhoff, more on construction and engineering.

More uses cases are expected for higher education, government and other verticals Unit4 focuses on in the coming months.

It is a nice example of “smarter work” as I continue my research on how automation is transforming various professions.

50(000) shades of automation

During my visit to Singapore couple of months ago I noticed taxis had front and back cameras recording live video. Like police body cameras they provide a digital audit trail. Why do taxis around the world, I wondered, not have similar with cheap cameras and storage costs? During a recent visit to New York city I noticed the toll showed up on the POS screen as soon as we crossed a bridge. Why do cabs in other cities, I wondered, not have similar integration between their toll tags and meters? I was admiring the 7 series BMW during a recent car service ride and the driver told me it was less than a year old. As he changed lanes I asked him why his blind spot sensors did not go on (BMW has had them as a feature for several years). He had never heard of the feature. I asked an Uber driver if we were tracking to the arrival time its app showed me.   He told me he uses his own GPS – hardly ever follows the guidance Uber provides him.

What’s my point? There are so many car models, local regulatory requirements, devices/sensors at play, can you ever see the taxi driver profession gravitating to one set of “best practices” any time soon?

As I continue my research on automation and impact in a wide variety of jobs I hear so many glib comments about one or two pieces of technology – process bots, 3D printing etc that are being positioned as automation nirvana.

Now let’s look at much more complex tasks. In agriculture and other fields such as pipeline monitoring, drones look very promising to monitor wide swaths of terrain. The drone platform itself is evolving nicely – in fact some complain there  are too many choices. The basic cameras are shrinking well to allow for space and payload considerations. But you need night-vision, infrared, multispectral sensors and other capabilities in the cameras and those continue to evolve. Machine learning to be able to find patterns in that much data for plant physiology analysis or corrosion analysis has to evolve even further.

What I am finding is available technology is for the most part assistive automation for pieces of jobs, not full blown replacements.

I know there is plenty of fear of job losses from automation. My fear is actually a bit different. As is common with so many technologies, we will go through another hype cycle and the benefits from automation will be over promised.

I winced when I saw Tom Davenport’s article that automation could bring in another wave of “reengineering”. If it is anything like the last wave, we could end up with even more labor intensive processes than we started.

Here’s a suggestion. If one of the usual suspects who did very well in the last wave of reengineering talks “automation”, ask them for a reference model for automating specific jobs. Start with my taxi driver example above. See if they have thought deeply about skills and tasks and the wide variations across countries and industries, and specific pieces of technologies to mirror and improve on those job skills. 

Next, ask them what they are doing internally to automate their own labor forces. If the answer to this question is unclear, I would be careful in hiring them to automate roles in your enterprise.

My belief continues that automation, applied well, leads to much smarter workers. The “applied well” part is something we all need to seek out and learn from. My research continues.

Workday: North by Northeast

I started the year with voices of customers still ringing in my ear from interviews for SAP Nation. They were fretting that there were few operational, industry-specific applications in the cloud. Workday and other cloud vendors were missing out on a window of opportunity around SAP and other on-premise vendors as they keep enriching back office functionality and targeting narrow markets like higher education.

Having spent the last few days with Workday executives and analysts who follow it at the annual Tech Summit and participating in the Predict and Prepare videocast I left feeling that that window of opportunity is still mostly open, and that in the back office, the Workday positioning has actually strengthened.

At the Tech Summit (a Twitter based transcript of which is nicely captured by Holger Mueller in Storify), we saw a confident Workday. The global footprint of functional and data center coverage is growing nicely - there was even a lucid discussion on the sticky Safe Harbor issues that multinationals face in Europe. The scale of customers being serviced is growing exponentially. Workday is investing heavily in lifecycle management tools, and keeping implementation costs in check, even as the ecosystem around it grows. The CIO of the large real estate services company, Cushman and Wakefield joined us all day and provided a customer pov to many of the analyst questions. Quite a symbol of confidence to have him participate and provide unvarnished input to analysts.

Predict and Prepare, the annual videocast that Bill Kutik hosts with a few analysts like me had its largest attendance ever this year. (Workday will have the replay up shortly). And in a sign of Workday's growing product portfolio and opportunities to cross-sell and up-sell, the majority of the attendees were incumbent customers, not prospects. The topics were far more strategic and focused on issues executives want to hear about including growth and automation impact on workforces. In other moves, Workday is marketing higher via golf and other sports to executive audiences and its executives are becoming more visible at Davos and other high level events.

In the meantime, SAP appears to have become even more exposed. S/4HANA, introduced earlier this year, is looking more like a Trojan Horse to sell HANA, not provide its application customers a a long overdue refresh. The initial implementations of the Finance applications (the only ones initially delivered) are almost entirely on-premise (more cynically "private cloud"). There is little visibility on when industry functionality will be available, and there is little integration between its finance functionality, that of HCM in SuccessFactors, or other procurement, T&E, contractor management functionality in other acquired features via Ariba, Concur and Fieldgass. Workday's cloud architecture and mostly homegrown features and integration across all these areas is so much more attractive.

Oracle has a more viable cloud back office story as Fusion continues to mature. But for all the valiant efforts of its cloud executives, the company continues to send mixed signals as it clings to its on-premise revenue and margins and annoys its customers with audits and its overreaction to support alternatives like Rimini Street. Similar to SAP,  Oracle has not moved much of its industry, operational functionality to the cloud.

Infor in fashion, hospitality and a few other verticals, Unit4 in municipal and higher ed have some traction in those markets. Microsoft, Plex, Netsuite have their own vertical strengths in retail and manufacturing but aim at a much smaller customer than Workday targets. So, the wider cloud industry opportunity has not shrunk in the last year. The window is still open, but will not stay so for ever, and I expect Workday to start to make moves in 2016 - cautious ones so as not to over stretch itself.

But if I was still at Gartner where I analyzed what it called "administrative applications" - for the back office of larger customers, I would put Workday in the top-right, leader quadrant. More emphatically, I would make it stand out in that quadrant, way to the Northeast corner.

2015 has been a very good year for Workday.

Workday Predict and Prepare–exciting new additions

Next Tuesday, the old pro (he has over a hundred such panels under his belt) Bill Kutik will moderate Workday’s annual look-ahead videocast.

There should be plenty of new wrinkles

Naomi Bloom and Brian Sommer, two of the best in their respective areas, will talk about rapid changes in the HCM and Finance functions.

Ray Wang is back and sure will work in his recent work on Digital Transformations.

I will share early results from my research on how automation is reshaping jobs in just about every industry and country.

We will be broadcasting from Workday’s brand spanking new studio.

And "the most interesting man in technology" Mr. Kutik tells us he has a few tricks up his sleeve.

So, if you have not yet, register here for this must see TV episode.

SAP Nation: a year later

I released Volume 1 of SAP Nation a year ago. It was a new genre for me. Previous books had been forward looking, focused on innovations. SAP Nation was more historical analysis and modeling to understand how we got to an economy as big as Ireland in such a short time frame. Volume 2 was more current as I focused on the S/4HANA launch but I also looked at 2 decades of other enterprise software launches and adoption patterns.

Here’s what I have found in the year since:

SAP Customers: Most customers I have talked to have thanked me for profiling so many of their peers in case studies. It has encouraged them to become more introspective. They have used it to review their own deployments and make adjustments.  I expect more to come out and freely laugh at the mirage of how “simple” things have become  (like the CIO of Dow who recently told the WSJ their 8 year implementation cost them over $ 1 billion). I expect others to say “no mas” (like DHL with a recent EU 350 million writeoff). But for the most part I expect customers to accelerate their diversification through ring fencing, two tiering, third party maintenance etc. By the time I wrote volume 2 in the summer I was impressed how many more examples I could see in each of those strategies.

SAP itself: Their reaction has been puzzling. They had heard quite a bit of what was in book directly from me for years before. I invited them to contribute to the book, and they did in small parts. For the most part, though this is a new SAP which is HANA, SuccessFactors, hybris, Jam focused. They appear to have little appreciation for the industry applications their predecessors had sold. They do not appear to have the curiosity to go through history – that the DHL write off is just the latest in a long series of failures that filled Chapter 7 in the first volume (even with DHL, SAP says “don’t blame our software”). Cannot blame them - even I had forgotten about many of those failures. What’s more interesting is this new generation of SAP employees keeps pushing their own products at customers, when the customers want to see value from earlier investments. Worse, they use audits and other tactics to try and squeeze even more out of the older products.  It’s boomeranging – I hear many SAP customers say it smells of desperation. Maybe S/4HANA was designed to reduce the burden of the earlier investments, but as I describe in the second volume I think the product will need to be radically rethought and re-launched.

SAP Partners: This is a mixed bag. Some are aligned with the new SAP and parrot how good HANA and SuccessFactors. are. Others keep clinging to the older economics even as cloud, mobile and other models are showing the need to change (and some of them have sent me nasty grams as if I created the cloud and other offerings they are being benchmarked against). The smarter ones are going in and replacing the incumbents with more efficient offerings. SAP’s own cost is only 10% of the nation’s GDP. The biggest changes have to come from its partners and they have a long way to go.

SAP competitors: Another mixed bag. The smart ones are not selling “rip and replace”(most do not have the broad functional or global footprint SAP does) but gradual transition. Some who have not even read the book happily recommend it to customers – not sure if it helps them unless they can show things will be dramatically different around their own products. Couple have told me they thought it is too nice to SAP. Seriously

Overall, I grew as an author. It made me research a lot more of the past. It forced me to listen to customers who were not necessarily happy (for books on innovation, the interviews are much more excited and animated). So, in many ways it was a very different experience.

But in a small way, I hope I am helping many SAP customers optimize their strategies.