Analytics & Data Mining

Is it that the machines are getting smarter … or that people are getting dumber?

(Yes America: I’m looking at you).

We’re worried that the machines might become self-aware.

Are we completely sure that, umm, people are self-aware?

Oh, and Yes: the accompanying picture was generated by a machine.

1. Analyse web-captured free text customer feedback to source good news stories and identify areas of risk or dissatisfaction
2. Analyse KPI results to determine which KPIs drive others, and which are lead indicators; use this to go behind traffic light indicators to unpack the *drivers* of performance
3. Benchmark offices or service outlets against each other to find best/worst performers; use this as a prompt to find the sources of high performance
4. Derive regional or area structure based on spatial distribution of clients
5. Derive whole-of-life cost estimates for hard assets/infrastructure. Model the impact on asset condition of various maintenance regimes over the expected life of the asset
6. Check for irregularities to supplement exception reporting and assure probity of accounts payable processes for internal audit, Auditor-General etc.

1. Analytics techniques can be used to 𝑠𝑖𝑚𝑢𝑙𝑎𝑡𝑒 𝑎𝑛𝑑 𝑝𝑟𝑒𝑑𝑖𝑐𝑡 𝑡ℎ𝑒 𝑟𝑎𝑛𝑔𝑒 𝑜𝑓 𝑝𝑜𝑠𝑠𝑖𝑏𝑙𝑒 𝑜𝑢𝑡𝑐𝑜𝑚𝑒𝑠 𝑓𝑜𝑟 𝑝𝑟𝑜𝑝𝑒𝑟𝑡𝑦 𝑑𝑒𝑣𝑒𝑙𝑜𝑝𝑚𝑒𝑛𝑡𝑠 𝑜𝑟 𝑙𝑎𝑛𝑑 𝑟𝑒𝑐𝑙𝑎𝑚𝑎𝑡𝑖𝑜𝑛 𝑒𝑓𝑓𝑜𝑟𝑡𝑠
2. Asset and facility managers can use analytics to model maintenance expenditure to 𝑠𝑡𝑟𝑖𝑘𝑒 𝑡ℎ𝑒 𝑏𝑒𝑠𝑡 𝑏𝑎𝑙𝑎𝑛𝑐𝑒 𝑜𝑓 𝑚𝑎𝑖𝑛𝑡𝑒𝑛𝑎𝑛𝑐𝑒 𝑐𝑜𝑠𝑡 𝑣𝑖𝑠-à-𝑣𝑖𝑠 𝑎𝑠𝑠𝑒𝑡 𝑐𝑜𝑛𝑑𝑖𝑡𝑖𝑜𝑛 𝑜𝑣𝑒𝑟 𝑡𝑖𝑚𝑒
3. Where large-scaled, detailed data for a given property market is available, analytics can be used to sift through the array of property attributes to potentially 𝑖𝑑𝑒𝑛𝑡𝑖𝑓𝑦 𝑢𝑛𝑑𝑒𝑟-𝑣𝑎𝑙𝑢𝑒𝑑 𝑝𝑟𝑜𝑝𝑒𝑟𝑡𝑖𝑒𝑠
4. For owners with multiple holdings, analytics can be used to help 𝑚𝑎𝑘𝑒 ‘𝑠𝑒𝑙𝑙-𝑣𝑒𝑟𝑠𝑢𝑠-𝑟𝑒𝑡𝑎𝑖𝑛’ 𝑑𝑒𝑐𝑖𝑠𝑖𝑜𝑛𝑠
5. Analytics techniques can be used to 𝑠𝑒𝑔𝑚𝑒𝑛𝑡 𝑏𝑢𝑦𝑒𝑟 𝑚𝑎𝑟𝑘𝑒𝑡𝑠 to break them down into discrete groups 𝑡𝑜 𝑤ℎ𝑖𝑐ℎ 𝑡𝑎𝑟𝑔𝑒𝑡𝑒𝑑 𝑎𝑝𝑝𝑟𝑜𝑎𝑐ℎ𝑒𝑠 𝑐𝑎𝑛 𝑏𝑒 𝑚𝑎𝑑𝑒 𝑏𝑦 𝑣𝑒𝑛𝑑𝑜𝑟𝑠 𝑜𝑟 𝑡ℎ𝑒𝑖𝑟 𝑎𝑔𝑒𝑛𝑡𝑠.