Mr. Hegarty’s Shared Services pilot: reality check

UT’s chief financial officer Mr. Hegarty has argued that a pilot should be run in order to assess the feasibility of shared services on campus. Given the professed goals of shared services this is either a deeply flawed proposal or a misleading description of the central administration’s actual plan. A pilot typically refers to a small-scale study done in order to assess the feasibility of and work out details in implementing a larger scale project or study.

Shared services gain efficiencies through economies of scale. The effectiveness of efficiency gains by such means can only be assessed through implementation on a fairly mass scale. According to a guide on shared services for management (Essentials of Shared Services by Bryan Bergeron, 2003)”There’s a minimum company size and revenue stream, below which the shared services model doesn’t make sense.” [1].

Furthermore, a drawback of shared services is high start-up costs, similar to starting a new company (one of the negatives of the model compared to outsourcing according to its advocates) [2].

Literally piloting shared services for a business would thus be a pointless and costly financial endeavor. One would have high-startup costs, and it would be impossible to assess the actual efficiency gains, since these are achieved through economies of scale.

It is likely, therefore, that the “pilot” is just a euphemism for initiation of the plan. Central UT administration has not spelt out any circumstances that would lead them to not follow through with the shared services plan in its entirety anyway. When Mr. Hegarty was asked about how the data from the pilot would be collected and how much this research would cost at the GSA on October 30th, he said he did not know. There appears to be no plan to assess the pilot, let alone assess it independently. Even if there was, it is unclear what these data would demonstrate because of how efficiencies are gained in shared services models.

These considerations make it understandable why UT central administration has made the pilot meetings for shared services closed to the public and campus at large.

[1] Bergeron, Bryan. 2003. Essentials of Shared Services. John Wiley & Sons.

[2] ibid

Shared Services at UT: The Accenture-Workday plan

Hegarty and Rohleder
(Steven Rohleder, COO of Accenture (left), Kevin Hegarty CFO of UT (middle)

The plan to implement Shared Services at UT was introduced by Stephen Rohleder, an executive of Accenture, in an op-ed piece in the Statesman [1]. Rohleder was appointed chair of the Committee on Business Productivity (henceforth CBP), which was tasked with giving recommendations to increase revenue at UT through more efficient use of its assets. This culminated in a report released January 2013, called “Smarter Systems for a Greater UT” [2]. The committee divides into three subcommittees; Asset Utilization, Technology Commercialization, and Administrative Services Transformation. It is the latter subcommittee, headed by Accenture COO Stephen James that recommends a shared Services Model that consolidates “Finance and Procurement, Human Resources, and Information Technology” for UT. These functions have typically been consolidated in the private sector. The CBP report states that its recommendation is based on research;

“The sub-committee gathered and analyzed a significant amount of data collaboratively with 27 different units within the University. This [sic] data strongly suggests [sic] that the recommendations are achievable based on the workforce distribution, the similarity of work functions, and the level of effort currently expended by individuals in each unit.”

Smarter Systems for a Greater UT, p. 19

Accenture was paid $960,000 of UT’s money in order to help the committee gather these data [3]. Key to interpreting and assessing the recommendation is how the subcommittee determined that the work functions were similar enough to make a Shared Services model feasible for UT. The campus community and its shared governance structures (Faculty Council, Staff Council, Graduate Student Assembly, Student Government) would be suited to assess this analysis since they perform and/or rely on these functions. However, the research mentioned in the quoted paragraph is still undisclosed despite multiple Open Records and FOIA requests. Skepticism regarding whether Accenture and the CBP subcommittee adequately took into account the specificity of department level functions remains.

UT released the “Shared Services Plan (Draft for Campus Discussion)” in October 2013. This draft contains estimations for costs of implementation and savings to be made through downsizing the workforce. Regarding the downsizing of the workforce the report states; “Total number of reduced positions ≈ 500 (≈4% of the total administrative workforce of ≈ 12,000 or ≈11% of the total administrative workforce of the HR, Finance, Procurement and IT workforce of ≈ 4,500) over 4 years” (Shared Services Plan: p.5).

The Shared Services Plan includes investment into the enterprise resource planning (ERP) software Workday (more on this later). The savings from downsizing to Shared Services are calculated as 280-320M in 10 years ($30-40M annually), with a $160-180M benefit. The investment costs are not itemized (change managers, programmers, ERP software, and other costs), but at least contain the costs of the ERP. Requests for an itemized breakdown of costs have been repeatedly denied.

The profits from the Shared Services Plan are erroneous or overstated. The plan claims that there will be a payback from the ERP implementation at Year 6. However, this calculation is based on assuming a $100M investment.  If the actual numbers are used there is no payback and we assume that 100 jobs are cut each year rather than 433 until 10 years after the beginning of the plan [4].

sharedservicesgraph1

sharedservicesgraph2

Another potential flaw with the profit projection presented in the Shared Services Plan is that it assumes that 433 jobs will be cut by December 2014. According to the plan “Attrition is expected to account for a significant percentage of the reductions”, we are not told what the significant percentage actually is.

As stated above, an Accenture COO was the chair of the subcommittee that drew up the Shared Services recommendation. Accenture was paid in order to facilitate the research on which this recommendation is based. There are three Accenture consultants on the Shared Services project team (Tim Mould, Ryan Oakes, Jamie Wills), including one ex-Accenture executive, head of IT at UT, Brad Englert [5]. The ERP Workday was bought in the summer with no campus feedback for a licensing fee of $30,000,000 [6]. Accenture conspicuously happens to be one of Workday’s most important deployment providers [7].

Thus, Accenture has acted simultaneously as consulting firm (in giving the Shared Services and possibly the ERP recommendation), and is involved and profiting from implementation of the plan (from continued consulting and programming fees). This situation has been identified as one of the primary causes of cost-overrun and poor performance in public projects. The dual-role played by the company presents a high risk of cost-escalation due to a lack of accountability in correct estimation and profit projections [8]. The faulty profit projections of the Shared Services draft Plan attest to this risk, even in the planning stages. The problem is accentuated when one considers Accenture’s history of fraud and corruption.

To conclude;

1)         Since the original data for the Shared Services recommendation have not been released and assessed by the campus it is far from proven that Shared Services is a feasible option for UT.

(2)        The profit projections have been proven to be erroneous and over-stated which either represents deceit on the part of Central UT administration or a predictable case of cost underestimation stemming from the structure of the plan itself.

(3)        The dual role of Accenture as consulting firm and as partly composing the project team runs to risk of continued cost-overrun.

In response to questions concerning the research done by the CBP on administrative services at a campus “town hall” meeting on October 30th, Mr. Hegarty claimed; “I didn’t want the committee’s information, the committee didn’t offer it and we didn’t want it.” [9] He was asked the same question at the Graduate Student Assembly and maintained that he was not interested in the data [10].

[1] http://www.statesman.com/news/news/opinion/rohleder-rethink-government-to-maximize-efficiency/nRcWk/

[2] http://www.utexas.edu/sites/default/files/files/UT_Committee_on_Business_Productivity_Final_Report_01292013.pdf

[3] http://www.utexas.edu/president/speeches/2013/smarter-systems-for-a-greater-ut

Participation by GSA officers in other campus based committees by GSA charged with taking up the recommendations of the CBP have found that this research support must have been given only to Transforming Administrative Services and Technology Commercialization subcommittees, but it is unclear how much on the money went to support each committee.

[4] Martinez, A. 2013. Should we centralize staff at UT? Presentation given to the Graduate Student Assembly, December 13th, 2013.

[5] http://www.utexas.edu/transforming-ut/committees/administrative-services/shared-services-committee

[6] http://www.utexas.edu/transforming-ut/committees/administrative-services/administrative-systems-committee

[7] http://www.accenture.com/us-en/Pages/service-acn-workday-alliance-video.aspx

[8] See Flyvbjerg,B. 2003. Megaprojects and risk: An Anatomy of Ambition. Cambridge University Press. and source cited therein.

[9] Podcast of the Townhall meeting; Hegarty makes that exact statement at approximately 39:00.

http://mediasite.aces.utexas.edu/UTMediasite/Play/b906091f708f416db977e7…

[10] For the GSA podcast please the following link. Look for the meeting on October 30th

http://www.livestream.com/utgsa