How artificial intelligence can revolutionize public procurement September 2018 @spendnetwork | @open_opps | @ianmakgill
Transparency Govts are mandating publication of tenders, contracts and spending. Transparency ensures fair competition and allows citizens to hold government to account. @spendnetwork | @open_opps | @ianmakgill
The data is useful We can already use the data to understand what we’re buying, how suppliers perform and where to source from. @spendnetwork | @open_opps | @ianmakgill
Single bids and avg days for suppliers to respond (UK)
Capita: Average revenue per customer, number of customers
4.5% 2017 increase in number of tenders for outsourcing: Outsourcing tenders 2017 increase in number of tenders for outsourcing: 4.5% @spendnetwork | @open_opps | @ianmakgill
0.7% 2017 fall in est. value of tenders for outsourcing: Outsourcing tenders 2017 fall in est. value of tenders for outsourcing: 0.7% @spendnetwork | @open_opps | @ianmakgill
The savings challenge The cost of savings is often too high, too much aggregation, contracts that are too long, too little competition. @spendnetwork | @open_opps | @ianmakgill
The savings challenge Suppliers will ‘bid to win, deliver to profit’. Planned savings disappear with change orders and poor contract management. @spendnetwork | @open_opps | @ianmakgill
The savings challenge Future savings will need to come from innovative sources of supply that drive competition, and better management of contracts. @spendnetwork | @open_opps | @ianmakgill
Exploiting data fully We can already use open data to understand what we’re buying, how suppliers perform and where to source from. @spendnetwork | @open_opps | @ianmakgill
Performance data We can also monitor suppliers and their contracts. Are they delivering on time and on budget? When do they fail to deliver? @spendnetwork | @open_opps | @ianmakgill
Continuously improve We can use data to show where we are going wrong, and seeing what delivers better outcomes. Economic development? We can prove what works. @spendnetwork | @open_opps | @ianmakgill
Savings from data Using data to find the right products from suppliers that deliver, at the best price throughout the contract. Anything that deviates from this is a false economy. @spendnetwork | @open_opps | @ianmakgill
New analysis techniques Having large, cohesive datasets (tens of thousands of documents) gives us the ability to spot and prevent problems early and make better decisions, using predictive analytics. @spendnetwork | @open_opps | @ianmakgill
New analysis techniques Descriptive analytics Diagnostic Predictive Prescriptive What happened? Why did it happen? What will happen? How can we make it happen?
New opportunities Predictive and prescriptive analytics change how we will work, instead of taking a “hit and hope” approach to tendering we can predict outcomes early. @spendnetwork | @open_opps | @ianmakgill
New opportunities Predicting which suppliers will deliver late or over budget or prescribing how to structure tenders that optimise the best pricing strategies. @spendnetwork | @open_opps | @ianmakgill
New opportunities Finding the optimum sourcing strategy. Using thousands of tenders to find the correct methods for getting value from a market. @spendnetwork | @open_opps | @ianmakgill
New opportunities Identifying non-competitive tenders before they’re published, showing where they might be improved. @spendnetwork | @open_opps | @ianmakgill
We need open data We need the data to know what is happening in markets and to use that as ‘learning’ data for algorithms. @spendnetwork | @open_opps | @ianmakgill
We need open data We need this data to be structured so that details of a contract from multiple buyers can be analysed in the same way. @spendnetwork | @open_opps | @ianmakgill
We need open data Open data keeps us honest, makes sure we don’t use these tools to restrict or distort markets. @spendnetwork | @open_opps | @ianmakgill
Like to know more? If you’d like to know more about predictive analytics and using data to improve procurement, come speak to me after. @ianmakgill @open_opps @spendnetwork | @open_opps | @ianmakgill