There was a time when government communication was relatively straightforward. There were ten daily newspapers, a couple of rolling news channels and a few other news or current affairs programmes to worry about and everything else was 'marketing', and therefore less sexy. The web? Funny thing no-one was very interested in, run (badly) by IT.
Then, some time around the turn of the century, some bright spark worked out that this web thing meant we could speak to the public. Directly. With no journalists to interpret what we said. Magic.
Unfortunately, we then discovered that putting stuff on the web didn't necessarily mean that anyone would read it. But the real problems started when they did start reading it. And blow me down if they didn't want to speak back to us.
Once that genie was out of the bottle there was no way of putting it back, and government has struggled ever since with the way in which it should interact with citizens through the super-accessible medium that is digital. And that's before we start talking about personalisation, location services, mobile ...
None of this should be taken as an indication that all is bad in the world of government digital. There are some incredibly talented people working there, and the recent hire of Mike Bracken from The Guardian - as executive director of digital efficiency and Reform Group - is a coup. DirectGov, government's best shot at a single citizen-facing portal, gets 30 million visits a month and satisfaction figures to envy, despite some grumbles.
The blueprint for the future is, of course, Martha Lane Fox's report on DirectGov, which I oversaw from the Cabinet Office. It is a fine report, which only suffers because the question asked was 'what do we do with DirectGov?' not 'how should government best use digital channels to engage citizens?' (Although she stretched the brief as far as she could.)
So it's not all broken. However, it is a struggle, so here are three thoughts about what we might expect to see from government in the next few years:
1. First, a continuing tension between the way government is organised around 20 departments and a citizen-centric view that seeks information, services or engagement from 'the Government'. Martha Lane Fox's report recommends a single government domain, and a team led by the inspirational Tom Loosemore has created alpha.gov.uk as a prototype.
The approach is revolutionary, as is the idea of making an alpha version public, but crucially, it hasn't yet involved asking departments to stop doing anything.
2. Second is another continuing tension, this time between citizens' primary interest in government web as a route to information and the ability to undertake transactions like getting a new tax disc, and government's wish for citizens to engage in developing and implementing policy. For example, if I come to DirectGov to report a crime, is it OK to ask me my opinion on elected police and crime commissioners?
Government is increasingly coming to terms with the fact that most online conversations about it do not happen on its web estate, and it must come up with a more sophisticated response than offering a minister to appear on Mumsnet from time to time.
All of which also raises a lot of questions about government's online persona. The anonymous 'government spokesperson' of print is hardly a good model.
3. Finally, the reductions in public spending will have to drive more services online.
The real question is whether that change is simply putting a web front end on a piece of broken bureaucracy, or designing a web process for the citizen that offers localisation, personalisation, access from any device and which is available wherever I go on the web, not just somewhere ending in gov.uk.