Microservices – What have we learned ?


Yesterday, I shared some of my thoughts on serverless architecture  and ended up getting a lot of feedback and a lot of it went back to SOA and then logically on to Microservices. So I thought it may be worth penning some thoughts on this topic as well. Microservices are not a fad – they get used quite a bit today across big important companies, although perhaps not very consistently. Also, I think we have enough experience by now to calibrate our expectations compared to a few years ago.

pexels-photo-302895.jpeg

What is microservices architecture?

I am sure there is an official definition out there somewhere. But a simple way to describe it is as a way to design a solution as a collection of services, each of which does one thing quite well independently. Each service is its own thing – own process, implementation, language etc . The collection of such services that solve an end-to-end problem is then orchestrated with a lot of automation capabilities to form an application.

Why is this a good thing compared to a “monolith” architecture ?

Separation of concerns is not a new idea in designing systems. Microservices architecture is built on this principle. The traditional way of building an application includes a front end ( html/js) , a database ( SQL/NoSQL/File/…) and App server to handle logic. When we hear people criticizing “monolith” apps – they are usually referring to the serverside logic built as a logical whole.

Monoliths are not bad per se – they can be designed and implemented in a modular way, and can scale with the help of good design using load balancers etc. Just that when it comes to scaling, testing etc – you have to deal with the whole even though only a small part needs to change. As cloud becomes more and more the default deployment option, the flexibility to scale and change quickly becomes a bigger need than in the past. Microservices is a very good way to deal with it. Many monolith systems will co-exist with the world of Microservices .

How micro is a microservice ?

This is one area where the wisdom from actual projects tend to throw cold water on the theory and philosophy of microservices. The tendency for many engineers is to go super granular in service definition . Almost without exception, everyone I know who started with this approach has regretted it and agreed to start with fewer services and then break them into smaller chunks over time. The operational overhead is quite significant as you play with tens of services – you now have to maintain and monitor several services, and at some point there is a performance penalty for too much communication across a bunch of services that all do one little thing.

Another interesting aspect is whether your system needs to behave more in a synchronous fashion or an asynchronous fashion. When you break the system into smaller chunks, essentially you are favoring asynchronous communication between them. Then if you need to make it work in a synchronous fashion – you may question your granularity decision quickly.

What about the product/project team?

I have seen several ways in which teams are organized , and have spoken to folks who worked in such teams where I had no direct involvement. There are a few consistent themes

  1. The need to communicate frequently and freely is a make or break criteria, way more than traditional approaches. With great flexibility comes great responsibility !
  2. One big advantage that comes with microservices is that each service can be implemented with a different fit for purpose language. And each service might choose a different database for persistence. While that is all great in theory, just because you can should not translate to you should. For large projects – too many technology choices leads to diminishing returns. Choose wisely !
  3. There is practically no good way to hand off to an ops team when dev is over. Microservices forces a DevOps culture – or at least DevOps tooling for sure. Its probably a good idea to get EVERYONE in the team some training in tooling. You need different muscles for this world than dealing with a Tomcat cluster. The promise of CI/CD needs a highly trained, high performing team. I may even venture to say that best practice is to have the same team that builds the team to continue to support and enhance systems built on microservices. There are just too many moving parts to effectively transition to a completely new team.
  4. There is no substitute for experience. There are not enough highly skilled folks around , so the ones you get need to carry the weight of mentoring their less experienced colleagues. Written standards might not be enough to overcome this. A common observation is two services looking at the same business object – like a vendor object that is of interest to an accounts payables service and a compliance service – and interpreting the semantics differently. Only with experience can you catch this early and converge.

Is it truly easy to make changes compared to monoliths ?

If you are a microservices fanatic, you probably are well versed in all backward compatibility tips and tricks, and hence your answer has to be YES. I will just say that there are some cases where you wish you were working in a Monolith, especially when faced with pressing timelines. A good example is the changes many apps will need due to GDPR  . When multiple services need new functionality – you need to wrestle with the best approach to get this done. Would you create a new service that others can call ? Maybe a common library? Maybe change each service and make local changes? Each has obvious penalties. No silver bullets – decisions taken in designing the app will dictate whether you buy aspirin from Walgreens sized box or Costco sized box 🙂

What about monitoring, testing, debugging etc ?

All the overheads on these topics that comes from distributed computing are in full force here. This is one area where the difference is significantly more noticeable than in the monolith world. Many of us are fans of doing Canary releases . You should have some consistent philosophy agreed on upfront for release management. Whether we admit it explicitly or not, lean and fast deployment has a tradeoff with testing effectiveness. Essentially you are relying more on your ability of montiring your app ( via all the services and messaging frameworks and redundancies) and making quick changes vs trusting impeccable test results . This is a significant change management issue for most technology teams and especially their managers.

So is microservices architecture a safe bet for future ?

There are plenty of public success stories today of microservices implementations – and conferences and tech magazine articles and youtube videos and all that. All Major SIs have expertise in implementing them. So in general, I think the answer is YES. However, I am not sure if microservices over time will be any less frustrating than monoliths in how they evolve. I probably will get some heat from my purist friends for saying this – but perhaps one way to smoothen the journey is to start with a monolith as we have done in the past, then as it evolves – perhaps have services that call the monolith’s APIs. And as you learn more, break down the monolith to a full set of services. I am not saying this because I am a non believer – I am basing it strictly on the talent available to do full justice to a pure microservices based architecture in a mainstream way. Just because Netflix did it does not mean everyone can. In any case – the mainstream pattern in large companies any ways is to start with their old monoliths and roughly follow the approach I mentioned.

Advertisements

SAP Needs A Better Cloud And Platform Story, And A Good Story Teller


This past week, there was a 2 day influencer event in Palo Alto where SAP explained (or tried to explain) its Cloud, Mobility and DB&T vision to about 16 of us. SAP paid for my airfare and hotel and ground transportation. Many thanks to Mike Prosceno and Stacey Fish – you are the best at what you do.

We started with cloud on Day 1 .  Cloud session was led by some of the smartest people I know in SAP, yet it did not go as well as I expected.  None of us in the room seemed to understand what the cloud story really is even after SAP tried a few times. As a technology fan, it did not matter to me that SAP had a sub par performance in the same week as Dreamforce was happening, but I am not sure if everyone else in the influencer gang look at it that way.

Sanjay Poonen and his team did the best job at the event when it came to explaining SAP strategy on mobility. I would say Sharron Ruddock, the COO for mobility, deserves special mention for  how well she explained mobility pricing. Steve Lucas and team took second spot – a close second I should say – with their HANA and DB&T story. And cloud brought up the rear.

If I travel back in time, when SAP execs were pitching ERP to customers, it was all about the business needs that ERP solves. It was not based on ABAP and Basis that a sale was made (which I felt bad about at the time, since I made a living as an ABAPer).  Fast forward to cloud era – and the whole conversation started with technology. Isn’t the whole idea about cloud that we don’t need to worry about technology as much as before?  SAP missed the boat on this one, and should try to revise the story line before their next attempt at Teched and SAPPHIRE.

It is not to say technology vision is less important – it is vastly important. But technology does not serve a purpose without tangible business benefits. If SAP has to succeed in cloud – it needs a VERY strong play as a platform.  This is also where that old DNA question comes to mind again. SFSF has the DNA for cloud based Apps without a doubt. But has that acquisition helped SAP inherit something useful for platforms ?  If it has, then someone needs to point it out, I did not see it.

There were a few highlights that definitely caught my attention. My friend Sethu, who is the deputy CTO at SAP, explained to things that are important.  The most important one is that SAP should not have multiple platform pitches – like mobility platform, analytics platform, Neo platform and so on. There should be  just one SAP platform going forward – which unfortunately does not have a name yet. Sad to say – this important memo did not reach the other parts of SAP in time. We got great explanations of mobility platform and Netweaver cloud platform right after Sethu told us there will only be one platform. The second thing that caught my attention from Sethu was that SAP is aiming for a unified semantic layer for everything that works on SAP platform. This is a much needed aspect of the platform story, and I am waiting to learn more about this.

SAP also has a good integration story that Nayaki Nayyar explained. I had met her once at Madrid last year, when we co-judged innojam before SAPPHIRENOW there. Cloud integration is critical to SAP’s cloud and platform future, and Nayaki explained that there is an integration technology they developed called Netweaver Cloud integration , and that it has two components – PI on demand, and Data services on demand.

The integration concept is webservices based. That is good – I get that, but that is not to say it completely reduces integration costs. It still needs harmonizing security, data, IDEs, and a consistent way to meter the services and components.

I am not fully convinced of the integration story on cloud. To begin with – I am shocked SAP positioned it as Netweaver brand. I don’t know a lot of clients who think of netweaver as a best in class integration platform. And calling its components as PI on Demand and Data Services on demand just gives the impression that SAP just repackaged something to make a new offering. Nayaki clarified that it is not so – and that it is all brand new.  I am sure I have said it before, and will say it again – without an acquisition story to go with it, I doubt SAP will have credibility in market with an integration story.  There are a variety of acquisitions possible based on what SAP can afford from TIBCO to snaplogic.

There is no MDM on the cloud from SAP in this whole cloud integration story. So essentially, one of the existing transaction systems will need to be treated as the master for the data that integrated systems will play with. I am sure this is just a matter of prioritization for SAP on where to invest for short term. While this is ok in convincing bloggers, I am not sure this is a good story to tell customers who typically need a longer term roadmap. Not having an MDM on cloud story makes it harder for them to buy into this integration strategy in my opinion.

I liked the analogy that Aaron Au used (he is the SFSF CTO, and a very friendly guy) of how Amazon evolved from “it is all about the shopping cart” to “shopping cart is just a small part, and only used to close a transaction” to explain how SAP is looking at User Experience.  It is probably the best explanation of Systems of Engagement concept that I have heard from SAP till date. What did not happen unfortunately is that SAP could not tie this example to their products and strategy in a coherent way.

Rainer Zinow was articulate as usual on the ByDesign side. I finally understood the thinking behind Financials on Demand. It is good to know it is not exactly carving out financials out of Bydesign. In fact nothing is carved out – just some parts are activated and others kept inactive. And the business case is mostly the ability to do Mergers and divestitures faster, by moving those entities to the cloud solution first.  That is a story I can live with and I see value in it. I did not hear a ByDesign for Large Enterprises story – maybe I missed it, or maybe there is no such thing. In any case, I think ByDesign and B1 ( Richard Duffy is an excellent speaker I might add)  did not get sufficient coverage in this event.  At least there were no surprises.

I don’t recall anything earth shattering in terms of bold new use cases for cloud. May be the intention was to set the foundation with this meeting, and make the big announcements at Teched or SAPPHIRE. Or maybe I missed out the transformational message when it was mentioned. In any case – there were heavy weight cloud experts in the room with me amongst the influencers, and one of them might explain that part better.

Sameer Patel could not present his vision for social and collaboration in the event proper, but he took time at lunch to educate me on that. I am duly impressed – and it is a grown up version of how social and collaboration should work in enterprises, and it puts it in the context of business processes where it belongs.  Sameer gets the prize for the best elevator pitch of the day.

There was a slide shown at the end of the meeting, but put under NDA – which is kind of sad, since that was the best part of the story. Maybe the next time around, they should start with that picture and work back to the technology issues and solutions. I know it sounded cryptic – but I respect SAP’s reason to put an embargo.

Bottom line – SAP should do some serious homework on unifying the theme of their cloud story . Knowing many of the people involved in this, I don’t have the slightest doubt that SAP will get its act together quickly. It is not to please us bloggers that they should do it – if they don’t do this in quick time , there won’t be much attention from their customers who are getting better messages from other vendors. HANA will run fine for SAP for another year with no issues – but without cloud and mobility picking up pace big time, SAP might not live up to big ambitions of its leadership team. And as much as the story is important, SAP needs to find a good candidate from inside the company or outside to say that story clearly to the world.

This was way longer than I expected when I started the rant here, and I am sure I missed a lot of points. But given my day job needs me to to run back to my customers 🙂

 

 

Cloud And Mobility Market In India – See It To Believe It !


As I am finishing up my visit to India, I cannot help but wonder why I chose to make a living in USA, and not here. As I stepped out the air conditioned room I was sitting in at my parents’ apartment to drive to a restaurant with my dad for lunch today afternoon – the answer became clear. It is the heat and  humidity, the pollution, lack of effective governance, inability of people to stand in a line, the terrible traffic and the lack of large number of dog shows.  It definitely is not about money any more – wherever I turn, I can see an opportunity to make a successful for-profit business.

Since I had some time on my hand, I walked around trying to gather some primary intelligence on what the potential opportunities are in India.

Any one who has been to India knows that Indians have a craze for mobile phones. I always thought that bandwidth is a big problem here.Guess I was wrong. I did not have a single dropped call here. I am typing this on a tablet using Wifi at my parents’ house in Trivandrum, while also streaming CNN on my PC . Speed is not the same as what I get in USA, but not bad at all.

My dad, who can use a computer – but usually stays away from it, has two. My mom has one. The guy who sells vegetables in the local market has three, the butcher has a belt around his waist that holds 4, plus the one he has connected to his ear via a blue tooth, the 9 year old daughter of our neighbor here has an iPhone and so on.  I am also (proudly?) convinced that my dad and his uncle can out-SMS most kids I know in USA.  All business is done on mobile – usually via SMS. My mom can book a table for dinner, call a cab, ask for home delivery, get a prescription refill –  all by SMS – and apparently so can everyone else I have met here in these last couple of weeks.

So I decided to ask a few of these vendors on how they manage this traffic of incoming calls and text messages. Apparently they are at their wit’s end on managing this. Most small vendors have paid assistants who answer the calls and text messages, and keep a tab on a piece of paper or an old diary and send back confirmation. The more advanced ones use a excel sheet.

In a given day, they lose about 20% of their orders (some as high as 50% on high volume days) due to clerical errors.  I know many of these people from the time I was a toddler, since I grew up here. And they know I do some work with computers.  Several of them asked me if I can write them something on a computer to help them do their business better. They did not press this issue since they all had a line of customers waiting behind me in most cases to do actual business. The only mobility scenario I found here that has a good degree of sophistication is mobile banking, and it is widely used.

This led me to the issue of how pervasive are computers and internet connections in this segment.  Although everyone in this “Very Small Business” category had one or more mobile phones – usually a smart phone, none of them had a computer at work.  Most of them said they have one at home that their kids use, and that they pay for broadband access. Sure enough they have no idea what their kids do with their PCs. So whatever solution these folks need – has to be done via a mobile device.

I would have died of shock if I found an Apple Store in Trivandrum – and I did not. However, I had a near shock experience when I found that there is a Samsung showroom exclusively for smartphones and tablets. I swung by the store, and it is pretty big and nice and carries everything that one would expect in similar stores in other countries.

Next up in my agenda was to move up the chain and talk to people who run bigger businesses – like builders, architects, automobile workshops, car dealerships , law firms etc. Again, most of them are folks I know from before, and/or known to my parents.

Their big problem is managing their financials, payroll and compliance without extreme trouble. On the high end, they have software – either commercially purchased, or built in-house.  They have in-house IT staff and an army of accountants to keep the business running and compliant. In the lower end – they have all bought a computer or two, and some accounting software, and an internet connection. But no one seems to use it well. Many of the smaller shops have not switched on the computer in years.

There is no POS integration to begin with, even in some of the bigger shops. And even in shops that say they use computers well – I could see the industrial strength printer working non-stop generating the big multi-column reports. None of the business owners knew with any accuracy on their working capital, or gross profit. Apparently they need to talk to their auditor’s staff periodically to see how the business is doing financially.  There is zero workflow that is automated. Workflow essentially is a bunch of people running around with print outs of emails, or shouting over cubicles.

There is apparently an e-governance initiative under way in the government. They are now keeping electronic copies of everything – but of course, they also print everything in triplicate and file the hardcopies just in case a disaster happens. It does not help that the laws have not kept up with computer advances – so some of this hard copy fascination is just a response to legal requirements.  Traffic tickets etc are still dispensed by hand, and I have not seen any officer on the street using a smart phone or something for work. Everyone has a personal phone – usually very advanced ones. May be government can save some money by encouraging BYOD here.

I had to renew my passport here, and walked into the local office. Before I left USA, I had uploaded all the required documents into their site.  The passport services is at least partially outsourced here. The lady from TCS who handled my case at first window took one look at her screen and said – ” Your file is no good. You have left many of the mandatory fields empty” . I honestly did not know what to say. I decided against asking her why her system saved my application if it had missing mandatory fields . Next up, she said “I see you have uploaded all the documents. Unfortunately, I cannot download any of them in our system. So you can just give me photocopies”. I felt smug that I had already anticipated this will happen. Eventually she handed me off from Window A to Window B. I physically took some paper work from A to B. Person at B told me I also needed to include my marriage certificate. I had it handy, but he would not take it – I have to restart at A . Ok, so I did that. Eventually I was handed over to B and then C . At C, the lady double checked everything that B did, and that was it – in 3 hrs, I was out of there.  I would love to meet the person or team that designed the process and the wonderful software, and get a copy of my paperwork autographed.

The last area I tried to understand was how sales force of these companies use IT, if at all. Field sales for these medium size companies have company issued phones, and in some cases smart phones. Some of them also have laptops. They file field reports either on paper, or in a word document that is emailed. Some one in back office then files it in appropriate folders etc. For things like price and availability , they use the company phone to check with their friends or with back office. T&E is all done with paper, and needs manual signatures for approvals.

In each of the above cases, there are existing solutions – usually in cloud as SaaS, and most of them have a mobile interface of some sort. Yet, I saw very little awareness – instead the smart people who run these businesses have just adapted their business model to overcome the lack of technical advances. I asked them if this was due to a limitation on their part to spend money. The answer was eye opening. Every one including the butcher with the belt full of mobile phones to the builder of big high rises is willing to spend money on getting a solution that will help their business. They are only too aware of what they are losing out.

I asked them if the local IT companies have ever approached them offering solutions. Overwhelmingly, the answer has been an emphatic NO.  Additionally, the perception I got was that the local companies – even smaller ones –  only care about winning work from abroad and executing in India, as opposed to winning work locally. I don’t know if this is because of their cost structures or for some other reason.

At least with these people I spoke with – there is some awareness of social media  but near zero awareness for social business. Some of them use Facebook to keep in touch with their children who live abroad. Only one person knew what twitter was. On the bright side, I showed some of them what to do with FB and twitter and some of these folks seemed to like it.

Finally, I did a gut check with some colleagues in Bangalore, and some old classmates from Trivandrum – apparently the IT companies in India do play heavily in domestic market, but focus almost solely on large enterprises in India.  From a couple of weeks of asking around, I am firmly convinced that cloud and mobility are both potentially big plays in India for the very small to medium sized firms.  At a minimum, I would urge my friends at software vendors all over the world to check out the market first hand . Seeing is believing.

Slightly off topic – I had to spend some time at a hospital here in Trivandrum this week, where my aunt was admitted. Absolutely the best doctors were in charge of the treatment for sure – and the staff followed absolutely the worst process ever engineered. Plenty of administrative “paper based” mistakes were made in the few hours I was there, and I almost had to pinch myself to check if I was in 2012 or 1712. This does not really need mobility or cloud to solve – just good old client server will do. Or even a better paper based approach – I just cannot imagine life and death issues being handled through the current pathetic process.

And Appleby says , It’s Cloudy out there…


For the first time ever on my blog, I have a guest post by my dear friend, John Appleby – fellow SAP mentor, blogger and an executive at Bluefin, in London, UK.  So without further ado – Ladies and Gentlemen,  I present to you, John Appleby 🙂

 

When you go to a major conference, tributes to the blogging gods must be paid. Not because we have to, but because the creative energy that surrounds such an event makes you think. And in this case I’d like to thank my good friend Vijay for hosting my thoughts.

And it was so that a senior executive in SAP came over to me whilst I was stuffing my face at lunch and said “John, if I understand you correctly, you believe it is still Cloudy?”. He was referring to their cloud strategy but I realised that actually, I have no clue what an Enterprise Cloud Strategy looks like.

It is true that there are two broad cloud business models that indisputably work.

Small and Medium Enterprises – the suite

I know that if I were starting a new business, I wouldn’t invest in IT assets. I would let each employee expense a laptop and provide then access to some cloud services. And it is so that Netsuite has grown to a $236m revenue over 12000 customers which means an average of $1600 per customer per month. Margins are low but customers are buying.

So as a small business, organisations may buy from a single vendor or from a collection – depending on what suits the needs best. Either way there is no desire to have expensive IT assets in an office somewhere.

Large Enterprises – innovation in LoB

Large Enterprise ERPs are inflexible and as a result, organisations like Salesforce, Workday and so on have poached customers from the large ERP vendors like SAP or Oracle. There are a small number of such potential solutions that are horizontal (i.e. apply to all industry verticals) and a very large number that are industry-specific.

Salesforce Automation, Procurement, Cash Management, Credit Management, Talent Management – all of these and more are relevant such areas.

Is there a demand for a Large Enterprise Cloud Suite?

I’ve spent several late nights discussing this with all manner of people, including good friend Dennis Howlett who claims “the suite always wins”. He might be right but I think that it is in conflict with the current cloud market dynamic.

The reason for this is because right now, it is a number of niche players that have thrown together an app, integration and have created massive demand. But the operating margins are very slim. For Salesforce it is 6%. For Success Factors, there is a small loss. The same for Ariba.

Most of those companies survive as far as they have by trying to grab land as fast as possible and by creating platforms that move quickly but are not carefully architected. So suppose you want some of this market – would you rather try and build something, or acquire?

The problem in building a platform

The challenge is simple: those organisations throwing together a platform can barely make the economics work. But if you build a full cloud platform then you lose agility and add cost. This makes no sense at all economically, and nor could you compete with the niche LoB players like Salesforce, that are normally bought directly by the business.

How do you convince Sales and HR that they should buy a common platform? What is the benefit of that? What’s more there is very limited benefit in doing Financials at all – it’s just a means for posting.

Is land grab the only strategy?

This leads us to what SAP have been doing so far. They acquired Successfactors for $3.3B and Ariba for $4.3B. Given that Salesforce and Workday are not for sale, there are not many horizontal vendors left. The major area is that of areas of Finance – cash management and credit management. Are there others? Other than that there are a bunch of LoB apps focussed on Industry Verticals.

But in the end, I suspect that grabbing users and apps is the only relevant strategy right now. What happens next is the interesting part. Common sense dictates that Dennis is right and the suite will win in time, and therefore the only sensible strategy has to be to build out a suite to underpin land grab.

Final Thoughts

One thing is for sure – I wouldn’t want to have Lars Dalgaard’s job. Getting this part right is near impossible. It’s cloudy out there.

SAP buys Ariba for $4.3B – rather pricey, but still a good deal


As I was fighting jet lag and lying wide awake on my hotel bed in Bangalore yesterday night, I saw the news on twitter that SAP is going to buy Ariba for $4.3B. When IBM bought Emptoris earlier, there was always a question of who will buy Ariba – and my bet was on Oracle. And I am happy that it is SAP who finally did it. I am not sure why SAP did not announce this at SAPPHIRE last week, but I guess it doesn’t matter really.

Jim Snabe told us at Orlando last week that he envisions SAP to be a big player in business networks. At that point, several acquisition scenarios flashed through my mind, but Ariba was not on that list. As a matter of fact, Ariba hadn’t even crossed my mind for a few years now. They make less than a $500M in revenue,so my first thought was $4.3B was too much of a premium. Now, looking at the market cap – it looks like SAP is paying about 20% above closing price. Given the high multiples companies like SAP are paying for buying cloud companies, I wonder when they will generate real profits.

Ariba’s Calderoni will be a terrific addition to SAP’s board. I absolutely loved his decision to hive off services wing of the company to Accenture some time ago, so that Ariba can focus on what it does best. He was a CFO before he got the big job, and he is on the board of directors of Juniper Networks and KLA-Tencor. SAP could definitely use his experience well.Of course Ariba was not always a cloud company. So there is a learning opportunity for SAP here on how they can transform their big on-premises model into a more cloud enabled organization, by using Ariba’s experience. The one thing I do not understand at the moment is how Lars will lead the whole cloud business unit, and Ariba remains independent. Hopefully some one can provide clarity on this.

It is a bold move from SAP’s side for sure. They just paid more than $3B for SFSF few months ago, and now shelling out another $4.3B for Ariba shows a firm commitment to cloud. And it is consistent with their stated strategy of focusing on customers, vendors, money and people.

As I think more about this – the contrast with SFSF is quite apparent. SFSF is a solution that addresses more people since it is for employees. However, it is only used occasionally during the year due to the nature of the problem it solves. Ariba on the other hand targets much fewer users – but the usage is pretty high since it keeps track of purchases globally. So, SAP now has expertise in both models of cloud solutions.

I have a feeling that my buddy Sameer Patel is excited at the prospects of this acquisition. Ariba, at the heart of what it does – is all about enterprise level collaboration for purchasing/trading, which is close to Sameer’s heart. It makes even more sense now that SAP wisely hired Sameer.

Unifying all these cloud solutions – some acquired, some home grown into one consistent architecture and UX, is a complex task. But who better than Vishal Sikka to take that on? I think Ariba will have tremendous use of HANA compared to SFSF. SAP also can derive some synergy with BW and BOBJ in the context of what Ariba does for enterprises.

SAP has its strength in some of the atomic processes in Source to Pay, and Ariba has in some. And there is plenty of overlap between SAP and Ariba for now – and I am sure that will get rationalized over time. With the combined force of SAP and Ariba, Source to Pay could see some real innovation (Source to Pay is one of the most frequently outsourced business process, mostly due to manual steps) in business process. Combined with HANA – I am sure we will see a lot of change in SAP’s and Ariba’s user base.

In general, but for the heavy premium they paid, I am a big fan of this acquisition by SAP. And If I am wrong – I will blame it on jetlag 🙂

Some random thoughts on SAP’s cloud strategy


As I am flying to Bangalore, I have been thinking of SAP’s cloud strategy – or at least what I understood as their strategy – a fair bit. In general, they did not blow me away. However, I do think they have a lot of things done right. So here are some random thoughts for what it is worth.

SAP is targetting 4 groups – Customers, Employees/People, Money and Vendors. I like that – that covers a lot of the enterprise.But if being everything to everyone was a successful strategy, Business By Design would have scaled greater heights by now, which obviously has not happened. So somethings need to be prioritized over others. The big players in cloud – like SFDC, targeted specific segments and did not try to be everything to everyone. And in general – including SFSF – they have attracted more customers and users than SAP ever did.

I have often raised the opinion that SAP should run cloud as a separate business so that they are not burdened financially or bureaucratically by the larger legacy organization. So I was thrilled to see that SAP allowed Lars to operate cloud as an independent business. And he brings tremendous energy and passion to the table, which is also something SAP could do well with. So organizationally – I think SAP is well set.  If there is one question in my mind on this front – it is whether there will be consistency of engineering across SAP anymore if cloud is a separate organization. But given Lars and Vishal seem to be best buddies, I guess it will work out just fine.

SAP will also  offer parts of BByD as loosely coupled apps – like finance for example. This might work out ok – except, I think Finance in itself is more of a necessary evil, than a true innovation. Without financial postings – the rest of SAP won’t work very well . I am not sure how many will buy the finance solution just to run their GL. A few will for sure – but I don’t see it as a growth engine yet.

SAP derives most of its money from the on-premises world. This poses two challenges. One – if SAP takes a suite approach on cloud, they will have to cannibalize some parts of on-premises solutions. And two – they have to integrate cloud with on-premises world. Neither is an easy challenge to overcome.

First about trying to reinvent the business suite on the cloud. Don’t get me wrong – I do think eventually suite needs to run on cloud. What I am not convinced is the need to do it now, since that is not a path of least resistance within or outside SAP. I think SAP would find it much easier to create apps that solve specific problems that on-prem cannot solve easily. For example – on-prem does Accounts Receivables very well, but not Collections and Disputes management because those users typically are not on-premises. So, instead of moving AR to cloud – why not just move collections and disputes management to cloud? And once that is successful – there will be very little reasons for resistance to move rest of AR, and eventually all of GL to cloud, even for big companies.

And about the loosely coupled thing – I am not exactly buying into the idea of “out of the box” integration of cloud to on-premises systems. One – SAP is not considered a leader when it comes to integration technology. And two – on premises world is heavily customized, so nothing really works out of the box there. I made a living as a programmer primarily because nothing integrates in standard. For SAP to have a credible integration story – they should probably buy TIBCO or something ( ok, sorry to my friends who have heard me say this a million times). But given the high valuation, it might not be feasible.

On the social aspects of cloud, I like what they are doing. Combining Jam and Streamwork makes total sense for customers. And I applaud SAP for their stance that “Business” is the better part of “Social business”.

HANA is a perfect DB/platform for cloud. It will be a safe bet for customers and SAP since HANA can evolve and mature much faster in an environment that SAP has full control over.

One last thought before I have to shut my PC off. What is SAP’s data center story? Cloud needs significant data center efficiency. I think I heard SFSF has its own data centers. If SAP is serious about cloud in a big way – I think it is in their best interests to tell the world more about their strategy on data centers. And to extend that thought – they might even have to buy a hardware company at some point.

I am out of here before the flight attendant of this AI flight will shut my laptop down for me herself. Let me know what you think.

SAPPHIRENOW 2012 Orlando – Day 2 – Cloud, HANA and Mobility


My initial idea was to write an update for each day after the session – but after the first day, that plan did not work out. So here I am finishing this post in India., while fighting jetlag. Day 2 of SAPPHIRENOW was a lot more chaotic for me than Day 1 for sure. The day started with Jim Snabe’s key note, including Lars Dalgaard’s presentation on cloud at the end.

Right off the bat – Lars brings a kind of energy to SAP that is usually only seen with Bill McDermott on stage. Lars came across as confident, positive and strong in my opinion. I was also mighty pleased to hear that SAP and SuccesFactors implemented all of each others cloud solutions internally in few months. I had a chance to meet Lars with my blogger buddies. He is an amazing guy with a laser like focus on details. Jon Reed did an outstanding job in that meeting, and came in well prepared with a comprehensive list of questions for Lars.

Here are some highlights of what we learned
1. Lars had a chance to kill Business By Design if he wanted to. But since he thinks the only missing part for the product to be succesful is a good go to market strategy, he decided to not kill it.
2. There is an opportunty for SI partners in cloud space, and he is looking forward to working with them.
3. He thinks HANA is the best thing that could have happened to cloud portfolio, and he will be using HANA extensively in his products.
4. Engineers from SAP and SFSF are being integrated now

As expected, Jim did a solid job on his keynote.It started with a video on SAP history. Funny part for me was Josh Greenbaum saying R/3 was an open platform or something such, and the guy sitting next to me watching the keynote choked on his coffee. So things started on a light note for me.

While BByD will continue to exist as a suite, parts of it will also be offered as smaller point solutions – like for Finance. I am on the fence on this strategy. SAP is a late entrant to the cloud market, and needs a strong differentiating message for its cloud portfolio. I think they could do one of two things – or even both.
1. Create horizontal solutions which cover large number of customers – for example, Collections and disputes management in cloud, with integration to BByD or On-premises FI-AR.
2. Vertical solutions for SAP’s industry solutions that cover the business partner network – like say dealers for automotive, or insurance agents for insurance companies.
I am looking forward to see how SAP differntiates in cloud going forward.

Jim also envisions the world will move in a few years to putting all the data in main memory. I beg to differ on this as well – I seriously doubt disk will get replaced soon. Later,I was happy to see Hasso’s vision is that of hot storage on main memory, and cold storage on disk . Of course, Jim could be totally right – and I am looking forward to seeing how this will unfold in the market with customers. I did a gut check with a few that were in the convention center – and none of them seemed to think all data will ever sit in main memory.

After the keynote, in a private blogger meeting – Jim discussed the future of HANA. BW on HANA is the next big thing on his mind to get traction in the market. I readily agree – since it is low risk and high reward for customers. Also – most BW systems are small, and SAP already has proven that HANA works with 100 TB, when the largest BW instance in the world is only 80 TB.

However, there is a big issue in making this work – and that is the sales enablement. BW has 17000 or so installations worldwide. If SAP has to hit 10% of the market in 2012 that Jim thinks they can hit – they need to close more than 10 deals every working day for the rest of the year. As I have mentioned many times before, the partner ecosystem has been largely shut out of HANA so far for many reasons. It is getting better now, but not to the extent that between SAP and Partners, 10 deals will be closed every day.

However, when we met with Bill McDermott – he did not define the challenge as 10% of existing market. His minimum goal is to double the revenue they got last year out of HANA and Mobility. That I think is an achievable target, and will not need 10% of the BW instal base to be converted. Bill did clearly mention that it is a bare minimum goal, and that he fully expects to beat that target.

Another difficulty with expediting the HANA sales process, according to SAP executives like Lucas, Enslin and Snabe, is the availability of hardware in quick time to do POCs. If SAP can put HANA on the cloud and ensure data security, this is a problem that will go away quickly. Customers can migrate HANA from cloud to a physical box, as HW gets delivered. Hacking the installer to put it on cloud is very easy – so hopefully this stops being an issue.

Jim did mention something that picked my interest a lot. His vision is an 8 day migration of a BW HANA system. Apparently SAP is working with some SI partners on migration factories where a customer can ship BW systems to be upgraded and sent back. I am not convinced this is workable in the field.

1. Most customers run a 3 system landscape for BW, and each will need to be upgraded.
2. Shipping them to India is not going to be easy due to security concerns on data
3. Very few of the 17000 are on 7.3 version which is needed for BW on HANA. It is not realistic to have an 8 day upgrade to both 7.3 and HANA.
4. Migration directly depends on quantity and quality of data in the system. If the system has huge data, and customer only has a short weekend window to migrate the data in production- it will be a challenge. Either they will have to archive a lot of old data, or they will stop migrating and just stand up a new system with just high value reporting.

But Jim’s vision is the right one to have despite all of this. If HANA is to become a volume play for SAP, they need to find a way to accelerate sales and implementations tremendously.

Another interesting couple of tid bits while we are on the topic of BW on HANA

1. John Appleby found out from one of his meetings that 3.x objects do not need to be converted for BW on HANA. They will work ok, but will not have the ability to take advantage of HANA. I did not know this till John told me on Tuesday. SAP could do better with communications.

2. How will BW (and eventually Business Suite) enhancements work in HANA? The enhancements are today done in ABAP server, and typically uses a row by row processing technique to process data, as opposed to set processing. So essentially they cannot make use of HANA’s power, and will become a bottleneck in performance. I asked this question to Hasso and Vishal on Wednesday afternoon, and they did not have a clear answer. Rewriting them manually into some kind of stored procedure is not easily accomplished due to volume of enhancements in a typical ABAP system. I am waiting for SAP to give us an answer to that.

We also spent considerable time with Sanjay Poonen, who heads D&T, Mobility and Apps. Sanjay is big on mobility ever since he took over that portfolio, and is very optimistic about SAP’s chances. He took our feedback to heart last time and was instrumental in getting developer access sorted out to a great extent. Afaria on cloud is a tremendous value proposition for SAP if the price point is really low. On Apps side, they do need to woo many more developers to make it work. But with open technology partnerships announced for Sencha etc – SAP has a shot at making money on mobility.

In my mind – for SAP to succeed in cloud and mobility, they have to immediately address one thing. That one thing – often ignored – is APIs. SAP needs to be good at APIs if it has to attract non-SAP developers to work on mobile and cloud solutions. And it is an easy target for the existing couple of millions of developers in SAP ecosystem. It is low hanging fruit in my opinion, and ripe for plucking. If I might go on a limb and make a suggestion to SAP – that would be to let Craig Cmehill form a team and evangelize that with the developer community. He is already a rock star with a big fan following, and will be readily able to hit the ground running.  Of course, some one at SAP needs to check with Craig first to see what he thinks of this 🙂

Day 2 was also a big party night – starting with Global Comms reception. After spending a few hours there, and catching up with several mentor buddies , I skipped over to the IBM party a few doors away. I was amazed to see how many customers were interested in Mobility and HANA and wanted to start projects on it. And then we walked back to Hilton at the convention center late at night, led by Mike Prosceno. There is an unconfirmed rumor that we passed the peabody hotel twice in the process !