So Similar Yet So Different: CA State IT Infrastructure vs. a Fast-Growing For-Profit Company
Chris Cruz, Chief Deputy Director of the California Department of Technology (CDT) and Deputy State Chief Information Officer (CIO), provided insight on sample initiatives he is currently leading. I was delighted to hear about the change he is driving and some of the innovative ways he is engaging with the public.
IT Architecture: Simplicity by design
Although a few of his points inspired a blog post, I would like to focus on his current action to move the state IT from an on-premise and decentralized format to Infrastructure as a Service (IaaS) and Platform as a Service (PaaS.) In 2016 the Harvard Business Review published an interesting piece asking “Is Collaboration the New Innovation?”(1)(2). It is a good read that touches upon collaboration in the form of a centralized data system. The 138 CA state departments, agencies, and commissions that had previously operated as alone-standing entities will now go through a centralized body that will streamline decisions. That may slow approvals down but imagine the potential created by the simple unification of data? Cruz excitedly shared that, starting next month, all state entities will be migrating to Amazon Web Services (AWS) and Microsoft Azure, which both include FedRAMP High services.
The benefits that Chris Cruz listed in his presentations were:
- Increased speed and agility
- Self-service
- Pay-as-you-go pricing
- Reliable, robust, and scalable
- Security and regulatory compliance (FedRAMP, CJIS, HIPAA, etc.)
The CA State leads with data security:
Our State CIO led with data security as the primary reason for this initiative. He stated that a centralized network will allow easier controls and protection of public data. The second reason he mentioned was cost savings through increased efficiency. The State of CA currently spends between $3B and $5B on IT Infrastructure. This initiative will allow efficiency gains freeing up to $3B that could be deployed elsewhere.
The automotive company I work at burgeoned from 600 to 40,000+ employees in the last 6 years. One of my roles was to scalably architect the Human Resources Information Systems (HRIS) during this time of hypergrowth. One challenge was exactly what Professor Baretto mentioned during his lecture: the company’s product was so incredibly innovative that I could not fathom allowing a lag in HRIS technology such as the one the CA state is facing.
Chris Cruz listed security as the number one reason during his talk, followed by cost. Working at a for-profit company, the efficiency gains (read cost savings) was my top driver. My second driver was user experience. The 40,000 employees, the executive team, and the board were my users.
The for-profit world leads with efficiency:
Over the years, I pushed the same initiative at my company that Chris Cruz is leading on the state level. My team spent years eliminating systems duplication, simplifying architecture, implementing secure mobile solutions, realizing volume savings through consolidations, and so on.
In an effort to scale faster, beyond the organic hyper-growth, my company started pursuing Mergers and Acquisitions (M&A.) Like many others, pursuing scale efficiency was top priority (Figure 1.) M&A activity is an attractive way to grow (Figure 2) but the investment only pays of if several things happen, among which successful post-merger integration. A Deloitte survey asked 1,000 executives what the top factor is for a successful M&A transaction. Effective integration was at 23% — the top in 2016 — and has maintained one of the top two spots in the past three surveys. “Information Technology is a major driver of M&A benefits, enabling a significant portion of identified cost synergies across the business.” (5) (Figure 3) My team had to decide if to let the acquired parts of the business exist as is or if to overwrite their infrastructure with ours. I experimented with both and, after a few deals, became a proponent of integrating deeply, and as soon as possible. The speed invited more risk of costly error but planning helped mitigate that (3)(4).
I have been reflecting on the two worlds: Merging core systems at an aggressively innovative for-profit company versus doing the same at a state non-profit level. This post is not a comparison but just a platform to exchange ideas. It was an incredible privilege to get visibility into what CA is doing and directly hear from the guy that cares about it the most.
Figure 1. What will be the purpose of your M&A deals for 2016?
Percentage of CFOs selecting each purpose (N=70.) Source: CFO signals, Jan 2016, US CFO Program, Deloitte
Figure 2: M&A Deals in Automotive by geography
Depth of green and size of text increase with number of acquisitions since 2006. Source: Deloitte Internal Analysis 2016
Figure 3: Drivers for operational deal synergies: IT
Source(s): Deloitte Consulting analysis of over 30 prior merger of equal transaction; Gartner IT Primer on Mergers and Acquisitions, Ansgar Schulte, February 2015
References:
- Is Collaboration The New Innovation? Harvard Business review (HBR). http://www.ey.com/Publication/vwLUAssets/ey-is-collaboration-the-new-innovation/$FILE/ey-is-collaboration-the-new-innovation.pdf
- Remix Strategy: The Three Laws of Business Combinations. HBR. Ben Gomes-Casseres, professor of international business at Brandeis University. https://hbr.org/product/remix-strategy-the-three-laws-of-business-combinations/12184-HBK-ENG
- Timing of integration: Should it be fast or slow. WSJ.
- The 10 steps to successful M&A integration. Bain & Company. http://www.bain.com/publications/articles/10-steps-to-successful-ma-integration.aspx
- M&A Making the deal work. Deloitte. file:///C:/Users/borya/Downloads/us-ma-making-the-deal-work-compendium.pdf
5 comments on “So Similar Yet So Different: CA State IT Infrastructure vs. a Fast-Growing For-Profit Company”
Comments are closed.
Boryana joined Tesla in 2011, at only ~600 employees, and helped the company through more than half a decade of hyper-growth. By early 2017 the company had grown over 5,000% and Boryana had built and led a number of teams such as People Analytics, HR Information Systems, HR Operation, Compensation, New Markets Readiness, Mergers and Acquisition Integration, and Business Operations.
Boryana was also the VP of Talent and Culture at Wikimedia, the foundation behind Wikipedia. At Wikimedia, Boryana oversaw all internal operations, people functions, workplace, and the team practices.
Following her passion of enabling and investing in human potential, Boryana also spends time as the Executive Director at the Straubel Foundation. The foundation's focus areas are sustainability, education and science.
A life-long learner and global nomad, Boryana has lived in Bulgaria, Germany, Austria, Russia, and the USA (since 2005.)
Thank you very much for writing this article. I now became really curious in which automotive company you have bee working that it has been growong from 600 to 40.000 in 6 years only. That is a tremendous amount! Wow!
And to speed up the growth M&A is definetely a nice way to do it. Can you tell me how difficult you see it from your perpective to integrate purchased companies in the own IT infrastructure? And how you e.g. even aquisitions have bee cancelled due to a difficult integration?
Users who have LIKED this comment:
Hello Jens, Thank you for reading the blog. I appreciate it.
I found it easy to integrate IT into our infrastructure but I think that I was extremely lucky because the CIO on top had done more than 70 integrations in his career. We agreed on crystal clear goals and agreed to move fast so that helped. The CIO was so incredibly good. Furthermore, he communicated so clearly that was easy to get buy in from everyone. We employed a phased approach: starting with bare bones in phase 1 (such as e-mail, calendar sync, connectivity, InfoSec, basic equipment, access.)
An M&A deal gets researched and vetted very intensely during the due diligence phase, which is before any announcements. There are many deals that don’t happen — In fact that is a very important part of the job — recommending to the exec team not do it and helping the company avoid a mistake. I had one instance where I recommended to pull the plug much later in the due diligence process (but still before integration.) That was unusual but the more data I collected, the more apparent it was that its the right decision. The reason in that case was poor culture alignment.
Thank you for your insights into the state of California’s IT transition to a centralized data system. One of the main questions I have is why the federal government does not use a centralized data system, and encourages the different states to start using one main data system. This cooperation could lead to many of the benefits that you described in your article, e.g. more secure, higher efficiency, lower cost. Many of the US states could learn from iniatives of other US states, and one of the best methods for states to learn from each other is to share data. It can help them prevent making mistakes that other states have made in the past, and try out policies that turned out to be effective in other states based on the data. Why do you think US states do not collectively use a central data system?
Users who have LIKED this comment:
Hi Teun, Thank you for your thought-provoking question. Not being a native to the US, I don’t know enough about the US Government to know if the reason is structural. It might be that states work independently for a certain structural reason. I hope someone that knows about the subject can chime in here.
I love your idea of a US data depository but wonder how long it would take the Government to make that happen? If there are more than 130 CA entities that need to be pulled together, imagine the country-wide number and how long it would take for all of them to agree to post their data. But just because its hard doesn’t mean that it shouldn’t be done. My guess is that its not operational complexity but something even more fundamental such as lack of interest to collaborate in that way.
Hi Boryana & Teun
One reason could be that these organizations all existed prior to computing, so as IT became more advanced the different agencies had to choose what to adopt separately and based on their own budgetary constraints.
Now that the option exists to centralize their data and systems, I’m sure it’s very appealing but as you pointed out would require dedicated time and budgetary resources.
Users who have LIKED this comment: