While there are “pockets of excellence” within the public sector when it comes to using data to improve customer experience, there is a lot to be learned from the private sector in the way it has done the necessary organisational reform required to back up technology change, says IBM’s Murray Bruce.
“People look for the quick wins,” he told The Mandarin, “but some of the things you need to do, they’ll never be quick. When you walk into the room and switch on the light, you know it’s going to come on. You’re relying on it every time you use it. They’re the things users don’t see, but if you don’t get those utility capabilities right, you’re never going to realise an agile front end.”
A common problem is implementing a technological solution on its own as a fix to a complex service problem, resulting in a flashy new IT system built over poor processes and/or poor data, argues Bruce, who is Director, Public Sector Asia Pacific, Japan and China at IBM’s Analytics Group.
“We’ve seen some agencies go and build a perfect utility, but there’s no users for it, and it’s a fail. Conversely, if they haven’t done the hard yards at the back end, the front end use cases end up becoming a rat’s nest. If the quality of the information in them is poor, users start turning away because it isn’t an effective way to use information.”
The comments follow a survey commissioned by Civil Service World, which found that 31% of British central government employees believed their organisation “was not using all available data to create a personalised journey for the customer”. 29% responded that their agency was not using unstructured data in real time to proactively meet customer needs.
To be able to build an effective “digital front office”, says Bruce. “You’ve got to link it in into your middle and back office.”
“A lot of the struggle for government and their reliance on data from other agencies is they don’t have a singular middle and back office and to do this — a lot of it is quite complex. But there are clear needs to do this to make things easier for citizens and cheaper for government,” he said.
There are some examples where government is heading down that path — but the complexity in most cases is not technological, it’s organisational. It’s encouraging to see the Australian government establishing the Digital Transformation Office to address this need, he says.
Integrated systems need to be a part of this change. “If you don’t have a good master data in the back end, you’re going to struggle. If there are ten different records of Joe Bloggs in your organisation, he’s not going to get the service he needs.”
Increasingly the data agencies need in order to understand the citizen “is not sitting in their data centre, or a structured data base. They’ve got to start drawing on data from across government as well as publicly available data and data in various forms – an example is tapping into social media. There’s a lot of publicly available data that can be brought together to help drive better and more targeted services for citizens,” he thinks.
‘That’s someone’s else’s job’
Public service organisations fail to use unstructured data effectively in part because they’re not set up to incentivise thinking in that way, thinks Associate Professor Helen Dickinson of the Melbourne School of Government.
At a recent event discussing big ideas on service delivery, Dickinson recalls that one participant from local government raised the issue of organisations not using data effectively to improve customer service.
“She said ‘we’ve got people in call centres who collect a huge amount of information and almost none of it gets used’. Only a bit gets used — the thing that’s being dealt with on that day,” Dickinson told The Mandarin.
“Everyone in that room shared exactly that sort of experience. It didn’t surprise me — public sector organisations effectively still think in terms of a combination of services and programs. Not in a blame way, they’re just not incentivised to think in that way.
“The issues here are because services are built up with no overarching narrative about their organisation, people tend to think about just their part. Staff are not incentivised to share information with other parts. It’s also hard to do that if you want to share information. There are data protection issues around sharing. The challenges can be quite substantial.”
There is no quick fix to encouraging collaboration, either. While some places have given specific people the role of facilitating collaboration across areas — giving efforts a clear focus — in the absence of wider cultural change this can just lead to the mentality of “that’s someone else’s job”.
One issue Dickinson sees is the problem that outsourcing can present to data sharing. There are challenges around how much information private companies and community organisations are willing to share in the context of competition. Using better contracting to decrease the amount of information kept commercial-in-confidence may help keep stakeholders better informed as to customer preferences.
Legacy contracts are another challenge, says Dickinson. In some cases contracts were negotiated by public servants who had not done a lot of contracting previously and lack the kinds of provisions that would allow for improvements in delivery. The result is that government has little choice other than to wait out the contract period and renegotiate a better deal.
One example she offers is from Birmingham council in the United Kingdom. One agreement, for a private company to deliver shared services, “had a whole series of stupid things in the contract, such as that the company got paid for each call picked up rather than for problems solved.” Although this created perverse incentives, the council was limited in its ability to break the contract due to the penalties it would incur.
“It’s becoming less of a problem as contracting in the public sector is becoming more sophisticated, says Dickinson. “People know to watch out for that now.”
‘A lot of the time hypotheses disappoint’
Data also needs to be used in smarter ways in the search for insights.
“The old approach to data would be, I have a hypothesis and if we analyse this data, we will find this, which we can take to the citizen,” says Bruce. “The new paradigm is you have all this data and you look for correlations. A lot of the time hypotheses disappoint after someone’s spent all this money on a use case and built all this capability.”
He recalls the work IBM did with a Waratahs Super 15 rugby team to find correlations between certain actions and how players train with, play and recover from injuries.
“We put GPS on them for a while and observed what they were doing after a game. Of all the data we collected, we came out with three powerful correlations that helped identify how players recover. We identified two players we thought should sit out an upcoming tour of South Africa and told the team’s management. They were surprised — they told us they had just had a call from the doctor to say those two players couldn’t go on tour!”
It’s now cheaper to store large amounts of data and the analytic software is improving, so the days of simply going into the data with one hypothesis are passing.
One example involved a government client, who would sometimes spend months or a year investigating customers for fraud before making a decision about whether to proceed with recovery or punitive action — action which would sometimes fail if the customer was unable to pay the money back. Improved processes, which give insight not only into known relationships between entities, but inferred relationships, have brought average processing time down from a year to a few days or weeks, offering a significant productivity increase.
Don’t leave it to the CIO
Bruce thinks that although there are some “pockets of excellence”, many public sector organisations don’t have the kind of support from senior executives — other than CIOs — to really push through comprehensive change.
“Other than on something inherently digital such as an e-health program, it’s hard to identify non-CIO digital champions. In the private sector, competitive demands and business imperatives have driven a savviness, that awareness, that is essential to overall transformation,” he said.
“This has transformed telcos, airlines, retailers, miners and the like. It isn’t the CIO you see leading these initiatives — it’s the COO, the CMO and other leaders who work closely to drive business outcomes with the support of the IT team. There hasn’t been that burning platform in government in the way that the competitive market environment has demanded within the private sector. We’re now starting to see that emerging non-CIO, CXO role driving the next wave of government digital transformation.”