Meet George and Judy Jetson.
They’re fictional characters who featured in the cartoon series The Jetsons, which aired on TV from 1962. It was set 100 years into the future. The family lives in a SkyPad Apartment, with a robot maid, and they drive to work and school in flying cars. They communicate via videoconference, and have smart watches. These are just some of technologies that the Jetsons predicted for 2062 that we have right now.
In the show, George complains of his heavy workload of pushing a single button on a computer as much as five times over three hours, three days a week.
Welcome to working in 2023
Fastforward (or rewind) to 2023, and technology acceleration and adoption has skyrocketed by the rapid diffusion of digitisation projects and AI models into everyday business applications. With the early majority now accepting that technology has changed business forever, what benefits, challenges and limits will technology bring to the future of bidding? In this article we’ll explore the use cases where technology is being applied, or has the potential to be applied, to revolutionise our profession. Like it or fear it – it’s coming.
1. Open data
With technology adoption moving faster in the last two years than the two decades before, automation has helped to speed up processes. But just as the internet became democratised, so too is open data and the way humans are unlocking it to work faster and make better decisions. Data is the foundation of digital, the key element for making more efficient and impactful decisions. Without data, technology simply can’t support human endeavour.
If you’re not convinced just how powerful data can be, let’s look at the procurement process during the pandemic.
When Covid-19 hit, the world simply couldn’t wait for a tender process and emergency legislation was put in place to bypass procurement procedures. A single Tweet from the NY Governor in March 2020 highlighted the desperate plea for supplies, and a footnote offering a premium price.
However without any competitive bidding or access to supplier data, there was no accountability. This led to price gouging; as well as contracts being awarded to family and friends. Worse still, states competed with each other for supplies; counterfeit products made their way into hospitals and sadly, medical supplies didn’t make their way at all.
A NEMXIS Survey, 2020 across 58 countries found that COVID-19 related deaths in every third country were due poor supplier selection. But the pandemic didn’t really create new problems for procurement; it simply highlighted hidden vulnerabilities and where it was broken.
If we could rewind time, how could this have played out if the systems were properly digitised, and what use cases could this have solved?
- What if we had a global database of pre-qualified PPE suppliers whose credentials were linked to third party verifiers so that quality standards, and track records could be searchable?
- What if suppliers were networked so that they could have collaborated and partnered to match supply to demand?
- And what if procurement departments were networked to reduce duplication the pressure on suppliers to spread themselves thin or forego opportunities?
“The future is already here – it’s just not evenly distributed” – William Gibson, author and originator of the term ‘cyberspace’
As we’ve learned, without the right systems in place, access to data, or human points of view – nothing better happens.
Ukraine as an example of data transparency and procurement reform
While the world was panicking over toilet paper, the Ukraine was able to source and despatch quality supplies in less than 48 hours. This is remarkable given only a few years earlier the country was plagued by corruption.
Procurement reform led to an amazing collaboration between citizens, industry, the government and data scientists when they open sourced their procurement data via the Prozorro system Using Prozorro, tender and contract award data would be openly published in a structured, standardised format using the Open Contracting Data Standards – another initiative that started with volunteers. This is significant because until now, not even the world bank or OECD has been able to get a consolidated accurate figure on global procurement spend.
Instead of it being just another portal, these standards provide a format that is machine readable. In the Ukraine, they’ve added business intelligence tools to their system so that data can also be analysed.
By having visibility of data and following the money there’s been real evidence of impact.
- 2.7 million dollars saved a day which they can channel back into essential services
- Accessibility to tenders has grown from 4% to 100%
- 96% suppliers are SMEs
- Based on the success of the model, private sector has now developed a B2B version.
Today, the Prozorro system is a great source of national pride and there’s incredible trust between the government and its people. Even in the midst of a war, Ukraine is quickly planning to rebuild. With open data, they’re able to find prequalified suppliers and are introducing selection criteria that includes meeting sustainable development goals so they can innovate by using smarter, cleaner technologies. The country is also planning to open up implementation data so that international aid and donors can see how they are progressing with the use of funds.
By challenging the status quo and making procurement transparent, the open contracting data standards have now been adopted by 30 governments worldwide.
For bidders this data is gold and data science can now help us to get better insights into:
- Buyer and supplier trends
- Seasonal demand to make resourcing decisions
- We can see contract prices and their duration; so we can keep an eye on expiry dates
- And we can also detect red flags or patterns in behaviour where the tender window is unusually short or the same supplier keeps being awarded – so we might make a no bid decision, or we might approach them to partner.
And if we flip this to feature our own internal bidding data there’s incredible insights we can also gather for competitive intelligence. If we can structure and classify our bidding data, we can codify it – and it can build a better picture of what’s happening over time.
We can extend this into risk assessments, and we can see correlations and the relational factors between our bids’ success and failure. We can even dive deeper to analyse the spread of proposal time and what’s causing bottlenecks.
Opening up data and making it transparent might prove to be a promising foundation for a procurement and bidding blockchain. To explain how this might work, imagine the blockchain to be like a subway map that contains tracks – or ledgers – with transactions linked through blocks. These tracks can be permissioned, or permissionless meaning they can be public or private (for example, the tender being issued is public; but the clarification question and answer period might be private). If anything changes, a new block is created and everyone is notified.
We’ve heard of blockchain being used in supply chain manufacturing – even artwork – to keep records to verify authenticity and ownership. But anything that has a lifecycle like procurement and bidding is a promising use case. So far, there’s been early pilots of procurement blockchain implementations in Columbia, Spain, South Korea and the US using blockchain for vendor selection, evaluation scoring and contract spend. And even though the World Bank has identified blockchain’s great potential by using the Open Contracting Data Standards as a foundation, it’s implementation might be a fair way off.
While the technology is here, and the theoretical benefits seem valid, it is quite a quantum leap for procurement departments and the many other stakeholders involved in procurement to build a consensus based ecosystem.
3. Semantic Search
The real potential having lots of data and opening it up across our organisations is that we can start to use smarter technology to make it more relevant. We’re getting used to this experience now with search engines and apps such as Spotify and Netflix making recommendations. As bidders, we’re always searching for information but finding it and knowing whether it is useful is where we sink a lot of time. Semantic search limitations (like Google) might help you to locate information, but it won’t necessarily give you the exact answer you need. This example could be likened to going to a doctor and asking for advice and being given a list of articles instead of a diagnosis.
Now, advances in semantic search combined with machine learning can give us a place to go to search for content, and behind the scenes allows the results to be more contextual. This kind of search or the reasoning behind it however, doesn’t ‘’magically’’ happen on its own. We need humans and domain knowledge to develop the data dictionary so that data can be mapped and the machine can read it and make sense of the relationship. We can apply this tech to bid management use cases like knowledge library search and Q&A pairing but from a strategic standpoint, it also holds potential for making recommendations around win/loss probability, bid/no bid decisions, and even risk based on previous outcomes using relational data to detect correlations. This is an exciting development.
4. Open AI GPT-3 / ChatGPT3
The tech world is currently going crazy over ChatGTP, an artificial intelligence program that can mimic human writing. As well as helping us find content and make strategic decisions, AI in bidding can assist with research and writing using natural language, not just keywords. This might make us feel both excited and threatened. Unless you’ve been living under a rock, most people have heard of Open AI’s GTP-3 and ChatGPT3. Launched in November 2022, it is a conversational chatbot built on Open AI’s GPT-3 technology (Generative Pretrained Transformer 3), a state-of-the-art language processing AI model developed by the research lab. The GPT-3 text-in text-out model was trained on internet text using Wikipedia and book content, giving it incredible creative capabilities.
Many of us unknowingly use it every day across hundreds of apps – from chatbots to predictive text and grammar checks. There are now hundreds of apps are being built on GTP-3 output, with Microsoft purchasing the exclusive license to its source code where it plans to incoporate it into its Bing search engine.
Now, GPT-3 has been opened up for developers to fine-tune on their own data, creating a custom version tailored to their application. Customizing makes GPT-3 reliable for a wider variety of use cases and makes running the model cheaper and faster. Two areas the model has proved to be strongest are its understanding of code and its ability to compress complicated matters. The latter makes it potentially useful for compressing long-form content into summaries for bid writers who are imposed strict word counts for each response. As the data gets trained and fine-tuned, bidding use cases could extend to knowledge search and content curation.
Where to from here?
As a bidding community, where can technology take us? That’s entirely a human decision, and the use cases we define and the boundaries we set. The danger of having machines reword our sentences is the allure of wanting it write the entire piece.
So instead of aiming for full automation, we should not put the technology first. Why not put humans first and aim for superhuman bidders? Technology that makes bid managers think better, become more aware, and makes the process of bidding more engaging?
Our digital future shouldn’t be to replace human intelligence, but to augment it by using tech to make us better at what we do.