I'll agree that the day-long interview isn't right for every office, but I suspect that if you peel back the covers, you'll find that regardless of the process and the interview formats, companies take a serious number of man-hours per candidate.
Why so much time?
In short - because hiring the wrong candidate is even MORE expensive. Unless you have a very easy firing process (some industries do!), you will be stuck with this person for far more than the few man-days it took to vet them. In a knowledge working industry, assuming the candidate isn't a very, very obvious mistake, you are probably up against:
2-3 weeks of non-productive time that is simply expected - they get briefed by HR, do all their forms, have to set up their system and deal with any typical computer access mess-ups. They have to learn something about what the team is doing, and then start doing it slowly, with lots of checking in.
2-3 weeks of uncertainty - if you have truly hired a bad fit, it can really take another 2-3 weeks to realize it. Particularly in a complicated job, people always have strengths and weaknesses. The job probably involves many types of decisions and actions and being bad at a few is OK, as long as the candidate is pretty good at most things. Collecting a general sense of "pretty good at most" vs. "pretty bad at too many" things is not something you come to in a day.
some decent interval of feedback and correction - most teams will give a bad guy the benefit to the doubt and try to correct his behavior. After all, an honest mistake is forgivable... the first time. The interval varies remarkably across culture, corporate culture, industry and type of work.
the firing process begins - some companies are getting clever and writing probation timelines into offers, to make the "this just isn't working out" process go much quicker. But in many places, an employee who's been on the job 6 weeks is treated the same as someone who's been on the job 6 years - formal warnings, discipline, 2 review cycles and so forth.
I'd say that even in knowledge work, you're looking at a low end of 8 weeks to get rid of a lame hire, and maybe even a year in some cases. And that doesn't count the wasting of other people's time. A bad new hire will also consume time from - management, the team, system administrators, HR and possibly many others.
Taking a net risk of around 70 days (low end) of lost productivity... spending 2-3 days on an interview cycle actually isn't bad.
It's a fair point that if you interview 10 candidates in day long interviews, your company has spent more like 20-30 days and the investment starts to look less wise. But I'd posit as well, that if the company is working with a day long interview format, they will do this for less than 10 candidates per position for the most part... unless their closure and review process is very inefficient.
What does the company learn from a day-long interview?
A lot. For a complex job there's a number of different factors to consider:
Technical - can the candidate do the job? Can he do it efficiently? Can he adjust and adapt as the work grows, changes and fluctuates due to the changing nature of business and technology?
Responsibility - will the candidate take ownership for the work? Follow the rules that must be followed? Be honest? Fit the company's model of ethical behavior?
Team skills - is he good to work with? Will he get on any body's nerves to the point of team dysfunction? Will he be the right mix of supportive and yet respectful of others to fit the team he's working in? Is another team or place in the company a better match?
Communication skills and other professional qualities - all that other stuff that fits with this job. Particularly crucial in that I've seen day-long interviews include simulated communication - presentations on technical topic, test problem solving with the team, etc.
Then there's all the people who have an opinion:
- The Boss
- The Team
- Other management - especially in a crucial position, a management position, a small company, or other vital cases, figure on several levels of management wanting an opportunity for an hour
- Human Resources - may seem like a waste of time, but usually HR has a vetting process that includes legal coverage of the company in certain aspects. There can be real value to what they do here.
And it becomes like quantum mechanics. Put the wrong boss in with the team on a team interview and the team won't ask the questions that need asking. HR really shouldn't do any sensitive vetting with others from the work side around. The bosses usually each demand their own hour... put 3 bosses, 1 team interview, 1 HR vetting process and 1 communication skills oriented demonstration, and you might as well feed the guy lunch, give him a tour and call it a day.
There's also something to be said for making at least a few key people be around the guy for more than hour. While everyone has varying levels of tolerance for putting up with annoyances from others - most people can put up with quite a bit and not even notice, if they only have to do so for an hour. I know, personally, that I can have a pleasant conversation, even with quite irritating people, for an hour. 4 hours, on the other hand, is a whole other ballgame. Unless I can actually get along and get some value out of the contact, I will be contemplating gnawing my own arm off to get away after about 2.5 hours...
Translating that to interviewing - minor irritants can get overlooked in an hour long interview. But extended contact will get key people thinking about how much they REALLY want to work with this person long term... which is the point.
What does the candidate learn from a day-long interview?
As important is - what does the candidate learn? Another worst-case scenario is that what seems like mutual attraction becomes a job placements that makes the new hire significantly unhappy. Training someone and accepting the 3-6 month learning curve on most complex jobs, only to have a candidate leave after 8 months is about the most expensive mistake a company can make. It may actually be more expensive than hiring a failure, since at least you KNOW the failure isn't someone you want to stick around.
So... if the candidate's future happiness with the company isn't a concern for the company - it should be.
Much of the interview process is, or should be, a chance for mutual discovery. After all the candidate should also feel he can work for/with - the Boss, the Team, The Big Bosses, the Key Stakeholders, and HR. Lack of trust or chemistry in any of these relationships can be a big contributor to leaving an otherwise as-advertised position. Similarly - the Technical, the team vibe, the Communication Expectations, the Corporate Culture - are all as important to the candidate - who has a career to think about - as they are to the company.
There's also a whole range of elements to the on-site day-long interview that are purely information points for the candidate:
What's the drive like? (if coming locally) - now is a good time to realize what parking, rush hour, commuter options, and local servicse will be. I'm actually rather amazed at how many candidates miss this - several times, in several companies I've seen candidates leave for issues relating to their commutes or the location of the site - even when they came for onsite interviews.
What's the space like? - While a recent contractor at my company joked that if you work in IT, you will expect to spend at least some of your career in a basement - it's good to know what the space is like. Be ware of any company that doesn't take you on a tour and show you where you and your team sit. For info sec and other high security jobs, it's also good to have a sense of what the security practices and procedures will be - it's not something a company will disclose completely, but if you just walked through a metal detector, figure it'll be there for the duration of your career and discard those enormous belt buckles now.
What's the vibe? - as much about the casual spaces as the work space. Does it have employee gathering spaces? Do employees actually gather there? You can bet if it's mod furniture and to perfect orchids, that's a "no". Look for the places with a decrepit coffee maker and mug-rings. What's the cafeteria like? Do people bring lunch? Are groups eating together? How about the team?
Some of this is knowing yourself - if you don't eat lunch with the gang, then maybe it doesn't matter if there's good nearby cuisine. But if you were counting on a friendly group that eats lunch together and goes for drinks regularly - then see what the options look like here.
Efficiency
I'll wrap this up with the thought that a decent part of an all day onsite process does need to be about making sure the interview team is efficient. An all day interview should be the final step. There should be a vetting process that weeds out all but a few candidates, and that all day experience should be about the final yes/no.
This takes a decent process and someone in management driving it. If you don't have a closure process where interviewers are polled and challenged to target the attributes that were a must-have, or a drop dead no, then the group won't learn enough from this process to make it worth the cost to productivity.
I have seen cases where I've thought that this final closure step wasn't properly executed. Much like becoming better at anything, it takes a critical eye and an honest interest in learning to get better at this, and with an ever changing business domain and technology spectrum, most managers and resource staffing people are going to have commit to learning a game that is always in changing.
In short - if the company doesn't tune the process, and all day interview cycle will, very likely, still be a waste of time. But do it thoughtfully and it can be a decent insurance policy against hiring duds.