Why You Need An In-Memory Action Plan

This is a third post of a series based on a SAP-sponsored breakfast meeting organized in Sydney earlier this year with speaker Donald Feinberg, Gartner VP and Distinguished Analyst explaining the “Nexus of Forces”: social, mobile, cloud and information.

In the first two posts, Donald covered why in-memory is disrupting everything, and why every organization will be running in-memory in 15 to 20 years time, and the business impacts of the new in-memory computing possibilities.

In this post, Donald explains why you should have an in-memory action plan — and how to create one.

These comments are based on my notes taken from the speech, formatted for legibility.

Why you need an in-memory action plan


You need to change the way you look at IT infrastructure, applications, and the infrastructure that’s running those applications. Truly, with some of these new technologies like in-memory technology, there are no barriers, things that you can’t do. Words like “no we can’t do it” start to go away.

I’m not going to tell you it’s going to be cheap, I’m not going to tell you there’s not going to be bumps in the road as you’re doing it, but things that you really thought were not possible are possible now. Period. [examples in previous post]

What do you do in your organization to start to adopt or use some of the in-memory technologies? You are going to spend money on this. Whether the TCO is less or not, you still have to build your skills, you still have to buy applications, you still have to buy the technology and infrastructure and things like that.


Build a business case first. Show the value of what you’re going to do. The return on investment may be long or may be short. We recommend short at first. Small projects with quick return on investment will get you more projects that are bigger and have a greater impact on the company. But you have to prove it first – that’s the key.

Assign a small team of people to look at this. Most companies don’t have a research and development organization in IT (the big ones do). But there’s no reason you can’t have one person looking at the things that are possible with the new technologies, looking at how they can make your current applications more efficient, or start to change how you use them.

So set up a CTO or department of the CTO that has somebody in there who’s just looking at the stuff that’s out five years or ten years from now, so that you will be ready to start to do projects with it when it matures to the level of risk that you’re willing to take.

Always do a POC, proof of concept. Do not just assume that because it looks good on paper it’s going to work for you. You need to test it with your data, with your applications, with your people.

Brainstorming. A lot of people don’t realize that your business unit people are much more IT-aware than they have ever been before. Brainstorm with them on what some of these things can happen, in the business, and how they can make use of it. Who has the budget today? IT? Or the business unit? So if you don’t do this, they’re going to do it anyway, and they’re going to implement the technology without IT. The big disadvantage is that the company doesn’t get the broad skill base that is necessary, and that technology is not shared across the business units. It’s much better to keep it in IT, not because you order it so, but because you are moving along in these new ways, with the business units and what they need.

If you believe what I’m telling you about in-memory technology, as being part of your future, it’s not too early to start to define a strategy for how in-memory is going to enter into your organization and be used.

You may decide that part of the strategy is “we’re going to wait two years to let it mature”. That’s fine, but start looking now at where it can fit and when within the organization, so that you’re prepared and ready to accept it when it comes along. If you’re an early adopter, start tomorrow. If you’re more risk-adverse, next year, the year after.

But at least understand the strategy for how this is going to fit in your organization, because as we believe, it IS coming, whether you want it or not, so you may as well start now to look at a strategy for where it’s going to fit in the future.

Questions and Answers

What would you reply to somebody who said “I’ve already got enough problems in my organization already”?

From a short-term standpoint, I can’t disagree with that.

But some of the new architectures and the in-memory technologies can maybe help you with some of the issues that you have today.

It depends on what the issues are. One of the issues a lot of people have is speed: my applications don’t run fast enough. So maybe there’s in-memory technology that can speed that up. Or maybe moving it to mobile will make it run faster.

Looking at the nexus of forces and looking at technology as a solution to some of your problems may actually help you short-term.

Cloud – maybe cloud can save you some time. I’ll give you a simple example: how are your development costs? Use the public cloud for that. Let your developers develop on an Amazon AWS.

Why is that good? Your people don’t have to set up the development environment. You make a phone call, and you have it. When the project’s done, you make another phone call, and not pay for it any more. You don’t have to go out and buy a server that then you’ve got to figure out what to do with after the development project is done. So there’s a place where cloud immediately can help you.

So some of this new technology is mature enough to solve some of your problems. And then, when you start putting your head together with the business units and start to have an impact on the competitiveness and the bottom line of the organization, that’s where you can really make a difference. Some of this technology may enable you, if you’re a retailer, to turn your inventory one more time a year. Is there any retailer that doesn’t want to do that? And not be out of something when somebody wants it?


If the business unit wants to be an early adopter, but the IT unit is risk-averse and conservative, how does the business user drive this change?

I’ve been around a while in this business. If there’s one thing I’ve heard over and over since I started in the 60s, it’s “IT has to communicate with the business”.  We’ve learned that lesson – that doesn’t work. Going out to dinner with your business liaison once a month and talking with them is nothing.

So one of the concepts that we came up with around twelve years ago, with respect to BI specifically, is the BI Competency Center. The reason that has worked is because it takes business people and IT people and puts them together, working together, not talking. So they make decisions together.

If I’m going to do a new project, all the business units decide what the priority project is. This is a concept that works. Some of your companies can’t afford to have full-time people in it, so you do it virtually: you have a meeting once a week. But they still manage projects, they still make buying decisions on products, they still set strategy for the company. The group should not be run by IT (which is hard to swallow sometimes) – but by the business unit. And most important: the CIO can not be the sponsor. It must be higher in the organization.

So if I’m going to have a “business technology competency center” where people from the business and the industry are going to get together to look at new technologies and where they may work, the sponsor has to be the CFO, the CEO, the Board, somebody like that. Then they will work together to do this.

Risk-adverse IT organizations are normal. You have a job to do to keep the lights on and you’re not going to do it if you take risks. It’s that simple – you’re not going to have a job if you take risks.

So how do you fit that with adopting new technology? Again, just like with the research and development with one person, you can take a couple of people from your organization as part of this “business innovation competency center”, sponsored by the CEO, so you can go hire some new people to do it if you need to, or move people over and backfill them.

They may take on a project with a business unit where you see tremendous value to the business, and you look at something that is, say, in beta. And you look at that technology to enable that business unit to be more competitive, more productive, more profitable, and it doesn’t affect the rest of your organization. You still can deliver the things you’re doing, because it’s “outside”. How do you get to that? You have to get senior management in the organization – not the IT organization – behind you. How do you get that? A small project, to demonstrate to them the value of this kind of thing.

Now one thing that comes to mind immediately: if you look at what’s happened in the past ten years with data warehousing – my area – every time there’s been a recession, database sales and data warehousing sales have dropped off. Except in the 2008/2009 worldwide recession, where every segment of IT was negative growth except DBMS, which was flat. In that environment, flat was positive.

Why? Because when the CIO came in to the CEO and said “I need more money to spend on my data warehouse” and the CEO says “are you nuts, with this economy?!”, you pointed to a flat screen on his wall that had key metrics of the business in “real-time” – for the first time, senior management, the CFO, COO, CEO, could physically see the value that information was bringing to their business.

If you can demonstrate physically to them the advantages of some new technology, then they’re going to buy into it and start to fund it. You can’t say something like “I want a new ERP package” – in an economy like 2009, that will get you fired for asking. But if you have some real strong value that you can demonstrate quickly or instantly to them, they’re going to spend money on it if they think it’s going to save money or help them. So that’s what you have to do. Lots of people say “only large companies can afford that” – but anybody can put it together with at couple of visionary people from the business units and one or two people from IT to put this together, and they can be virtual.

Pfizer is one of our BICC case studies. They have 150 people full time in the BI competency center: 75 employees and 75 consultants. Most people can’t afford to do that, and I’m not suggesting you do. But here are models in-between that make sense, that will fit in everybody’s budget.


2 responses to “Why You Need An In-Memory Action Plan”

  1. […] In this post, Donald explains the business impacts of the new in-memory computing possibilities, and in the next post, how to create an in-memory action plan. […]