REQUERIMIENTOS GENERALES

From our analysis of big data in five domains, we reached a sense of what characteristics indicate higher or lower value potential from the use of big data to capture value from its use, as well as higher or lower barriers in realizing that value in different sectors. Using these insights, we created two indices: (1) an index on value potential, and (2) an index on the ease of capture. Each of these indices comprises multiple criteria which give us a relative sense of which sectors may be poised for greater gains and which sectors would face the toughest barriers. We do not claim that these indices give a full picture, but we believe that they give a good sense of both the potential value available and the ease or otherwise of its capture across sectors.

VALUE POTENTIAL INDEX

The five criteria we use in this index act as a proxy for how well a sector can benefit from one of the five transformative opportunities we have discussed in this report (Exhibit A1):

1. Amount of data per firm. The larger the amount of data per firm, the more it indicates that a firm is likely to be able to benefit from increasing transparency in terms of data. We used the storage available per firm as a proxy. We built upon our data mapping analysis to estimate the available data storage, in bytes, in 2009 in each sector in the United States. We then divided that by the number of firms with more than 1,000 employees (to avoid skewing the numbers by the large number of small businesses and individual proprietors).

2. Variability in performance. The higher the variability, the more it indicates a firm can benefit from the use of data and experimentation to expose variability and improve performance. We used the variability in EBITDA (earnings before interest tax depreciation and amortization) as a proxy. Within each sector, we took the EBITDA of major companies (with greater than $500 million in sales) from 2002 to 2007 and identified the 10th and 90th percentile EBITDA. The difference between the top and bottom performers became the variability we measured.

3. Customer and supplier intensity. The more customers and suppliers a firm has, the greater its potential to apply segmentation to tailor courses of action. We used the number of frontline employees (defined as those who interface with customers or suppliers) per firm as a proxy. We used data from the US Bureau of Labor Statistics to identify the number of frontline employees (major occupations include Standard Occupation Classification codes 41 such as sales clerks and agents, and 43 such as administrative workers) in the latest year available. We then divided by the number of firms with more than 1,000 employees

4. Transaction intensity. The higher the transaction intensity, the more likely the sector can benefit from the use of automated algorithms to augment or replace

human decision making. We used the amount of processing power of an average firm in a sector as a proxy. To arrive at a relative sense of processing power, we used capital stock data for PCs and mainframes by sector from the US Bureau of Economic Analysis and divided by the number of firms with more than 1,000 employees in each sector.

5. Turbulence. Turbulence, or how frequently leaders and laggards in a sector change place, is a proxy for the amount of innovative disruptions to which a sector is susceptible. We hypothesize that the higher the turbulence, the greater the likelihood that a sector will benefit from the use of big data to innovate business models, products, and services. Within each sector, we calculated the turnover percentage—the number of new companies placed in the top 20 ranking in 2011 compared with 2006, divided by 20.

Once we quantified each criterion (the proxy), we gave each sector a score of from one to five based on the quintile into which it falls into for each criterion. The overall value potential index is the average of the scores across the five criteria.

Exhibit A1

MGI has compiled a heat map of the value potential of using big data across sectors

Cate- gories Sectors

SOURCE: McKinsey Global Institute analysis

1 See appendix for detailed definitions and metrics used for each of the criteria.

Top quintile (highest potential) 2nd quintile 3rd quintile 4th quintile Bottom quintile (lowest potential) No data available Manufacturing Construction Natural resources

Computer and electronic products Real estate, rental, and leasing Wholesale trade

Information

Transportation and warehousing Retail trade

Administrative, support, waste management, and remediation services

Accommodation and food services Other services (except public administration) Arts, entertainment, and recreation Finance and Insurance

Professional, scientific, and technical services Management of companies and enterprises Government

Educational services Health care and social assistance Utilities Goods Services Re gu la te d and public Overall value potential index1 Amount of data per firm Variabili- ty in per- formance Customer and supplier intensity Trans- action intensity Turbu-lence

EASE OF CAPTURE INDEX

This index is made up of four criteria, each of which aligns with a key barrier to the use of big data that we have identified (Exhibit A2):

1. Talent. The more deep analytical talent a firm has, the better a position it is in to realize value from big data. We divided the number of deep analytical talent in 2008 by the number of firms with more than 1,000 employees in each sector.

2. IT intensity. The more IT assets a sector has on average, the lower the technology barriers to be overcome. We calculated IT stock using data from the US Bureau of Economic Analysis and divided that total by the number of firms with more than 1,000 employees in each sector.

3. Data-driven mind-set. This indicates how receptive the organization is to using big data to create value. We leveraged the latest survey results conducted on IT strategy by McKinsey’s Business Technology Office, which asked leaders the degree to which their organizations make decisions based on experience and opinions or based on data.

4. Data availability. We use the relative number of databases related to each sector in a proprietary corpus of data as a proxy for how accessible data is in a sector. Again, once we quantified each criterion (the proxy), we gave each sector a score of one to five based on the quintile into which it falls for each criterion. The overall ease of capture index is the average of the scores across the four criteria.

Exhibit A2

How MGI’s estimate of the size of big data compares with previous external estimates

What was measured Amount of data

MGI storage-based approach IDC/EMC1_Digital Universe UCSD Hilbert, López

▪ New data stored in enterprise external disk storage in a year

▪ New data stored by consumers in a year

▪ 7.4 x 1018 _bytes

(includes replicas)

▪ 6.8 x 1018 _bytes

Year of estimate

▪ For 2010

▪ Annual digital data captured (includes all generated, stored or not)

▪ Includes more than 60 types of devices

▪ Did not include information consumption by users through TV, video gaming1

▪ ~800 x 1018 _{bytes ▪ For 2009}

▪ Includes both digital and analog data for TV, radio, phone, print, computer, comp. games, movies, recorded music, etc.

▪ Measured data from consumption perspective2

▪ 3.6 x 1021 _{bytes (total}

consumption US only) ▪ For 2008

▪ 24.5 x 1018_bytes ▪ 6.49 x 1018_bytes ▪ 32.5 x 1018_bytes ▪ 123 x 1018_bytes ▪ 276 x 1018_bytes ▪ For 2007

▪ Capacities for specific technologies

– Server and mainframe hard disks

– Other hard disks

– Digital tape

– PC hard disk

▪ Total digital storage capacity

1 Includes chip cards, floppy disks, camera, video games, mobiles, memory cards, media players, CDs, DVDs, Blue Ray disks, PC and server hard disks.

2 Consumption is defined as the data each time used by the user.

SOURCE: IDC white papers on Digital Universe, sponsored by EMC; Bohn and Short, How Much Information? 2009: Report on

American Consumers, January 2010; Hilbert and López, “The world’s technological capacity to store, communicate,