Myth of the Flat Fee | Hacker News Reader

Myth of the Flat Fee

80beans.com

56 pointsroytomeij30 comments

Of course this is a good message to send to big clients with big projects who merely "don't get it". On the other hand, there are a lot of small businesses and other clients with very limited means who have maybe $1000-$10000 for a project. Web development is new territory for them, and they have no other way to approach the problem other than to put a fixed-budget stake in the ground, and of course their requirements are going to evolve as they learn.

One of the reasons I joined a startup was because I was sick of working under those conditions, and writing an FU blog post like this is the luxury afforded to those of us who can more or less pick what we want to work on.

That's great, but there is still a pool of small clients who can be profitable, but you have to understand how to work with them:

- Learn to use off-the-shelf software to build lots of functionality quickly in a predictable fashion (WordPress, Drupal, Expression Engine).

- Charge a flat fee only for something you've done before, and make sure it has padding. It's not unethical (as suggested by another commenter) because this is how you are able to afford a flat fee. You'll be churning out hundreds of these so you need to work the law of averages. This allows you to be a bit accommodating instead of going into conflict mode every single time.

- Turn your creativity to your own systems. Face it, on a $5000 budget you are not going to do anything revolutionary for a client. You can however build your own systems that you leverage in order to serve a hundred clients better. Do it well enough and you actually build a business that scales better than top creative agencies working on 6-figure budgets.

- Be customer-service oriented. Face it, someone paying you $1000 a year is just not very important to you and it's easy to let that attitude show. The thing I've found is that small clients actually are more accommodating than big clients as long as treat them with respect. That said...

- Dump toxic clients. There are always those with unreasonable expectations and no respect for what you do. The worst is when they are sort of borderline stringing you along, but you end up losing time and money over the long haul. How you get rid of them is a matter of tact; maybe you say you are all booked, maybe you raise your rates for them to an exhorbitant level, maybe you just give it to them straight ("I am not profiting from having you as a client"). I don't have a one-size-fits-all solution, but get rid of them you must.

  Learn to use off-the-shelf software to build lots of
  functionality quickly in a predictable fashion 
  (WordPress, Drupal, Expression Engine)

Bravo. This is the way I look at it: if it has been done before, somebody, somewhere has productized it. If it isn't a full-blown CMS, it's a bundle of Rails plugins and all I'm doing is writing glue code.

If it hasn't been done before, because it's really specific to a client's domain and idiosyncratic processes, it's R&D and we shouldn't pretend we know how to do it with enough rigor to provide a competitive quote.

In many cases, the problem is that it has been done before, buut the client doesn't want to "colour within the lines" and run with the limitations of an existing CMS or other platform, they want a bunch of customizations that have little or no ROI.

The Big Sell in those cases is convincing them to scale back their expectations about customization and live with having the 20% of the features baked into the off-the-shelf system that deliver 80% of the ROI. Even if they want to pay by the hour, it's not a good investment to build what amounts to business chrome.

Interesting tangent to this, after doing PHP from 2000 to 2005, I hopped on the Rails bandwagon full-bore and was using it even for small clients. 6 years later I feel that it was a bit self-serving because those old Rails sites are creaky and painful to upgrade—or even to find someone else to work on—compared to my PHP sites from the same era.

Now that I have ample outlet for my creativity I can be a bit more objective and admit that as much as I love Rails, it's a terrible platform for build-it-and-forget-it. Rails shines for core business apps under continuous evolution.

What do you consider to be a good "build-it-and-forget-it" platform?

PHP, Java, Cobol? It's all relative I suppose.

You're recommending COBOL as more maintainable than Rails? I don't know Rails, but I find this very hard to believe.

The times, they are a-changing. COBOL has its own agile framework:

http://www.coboloncogs.org/INDEX.HTM

Where did I say maintainable? This is about stability over time.

>creaky and painful to upgrade—or even to find someone else to work on

What do you mean by maintenance, if not that?

All great tips, thanks for sharing. I didn't put a disclaimer in the post, but our experiences are indeed based on larger projects, typically at least in the $20,000+ range. Those are always tailor made apps, nothing off-the-shelf (apart from the framework and all the open source components we use, but nothing like Wordpress). What we do is often pretty specific.

I think your comment, as you said, has some great insights when it comes to smaller projects.

It's not unethical (as suggested by another commenter) because this is how you are able to afford a flat fee.

I agree and want to expand a bit. Even if a change or addition only takes me 10 minutes of my time now, it's important to remember the months and often years of experience that I'm drawing upon to make that change so quickly. Explain to the client that they are not only paying you for your exact time, but also for your skill set and ability to do a job quickly AND correctly. All of those qualities cost money and clients who don't want to recognize that fact are toxic and should be dumped (as suggested by the poster I'm responding to).

That's right. In flat rate contracts, the client is paying for the value being delivered, not time.

I've got myself stuck in a "fixed price" contract, and while the number of features the client's asked for haven't really increased all that much, the depth of interrelation between those features was not at all apparent up front. Throw in agreeing to migrate existing data in to a new data structure without really understanding the nature and complexity of the existing data, and you've got a recipe for a long-overdue project. No one's happy. :/

Oh yeah, you can easily bill a month of full-time work for fixing an old app with a busted data model. After that you reach the place where you can sell a change that takes 30 minutes of work for $150 and really do it in 30 minutes, but getting there is expensive...

I'm at the point now where the time invested in the project has paid off - I understand the business problems much better now, and some of the intricate work I did months ago is paying off in spades (basically, having put together a flexible data model up front). Things that might have taken days to do before I can do and redo in minutes or hours. But it's taken way too long to get here, and ultimately this is a one-off - most of the knowledge gained isn't transferable to any other projects directly.

I've been there and I know how that sucks. In those cases I figured it was best to take my loss and get it done and over with as soon as possible (I'm sure that's what you're doing).

In my experience there's one thing you need to watch out for, and that's being a nice guy. You might start feeling sorry for your client because stuff is taking longer than planned, which makes it harder to say no to just minor changes to the scope. They will only take you 5 minutes each, and it keeps the client of your back. Those are the things that really get to you when it comes to staying motivated I think.

yes it is. I'd budgeted for 4-5 months of time, figuring it might go to 5-6. I'm in month 9, and it will be at least another month. :/ Huge lesson learned, and hopefully this will be the last time I learn it. I don't typically take on projects this large, and in the past if there was a 80% time estimation error, the impact wouldn't have been as dramatic. :/

A few large consulting companies in the 90's, Cambridge Technology Partners and Viant, for example, were very successful with this model. They were selling against the $100-200/hr Big 6 type consulting firm so there was plenty of room budget for the "risk" associated with going over the time estimate. The other key to their success was getting the client to agree to pay for an upfront 'spec'ing' phase before bidding on the development work. If you do the spec'ing phase correctly you can give a fairly good bid on development and truly assess what is in scope and what is a change order.

"Lies, damned lies, and software development estimates:"

http://weblog.raganwald.com/2007/06/which-theory-first-evide...

The flat fee game the OP describes is pretty much standard operating procedure even for big consulting companies like IBM and Accenture. First they agree to a flat fee (sometimes after a bit of requirements exploration) that cuts their margin for error fairly close. Then, after the work has gone down the road a bit, they bury the client under a blizzard of change requests for every last detail that wasn't documented in the initial requirements. The change requests are where the real money is made for consulting companies, especially if they're getting their employees to work unpaid overtime to fulfill the extra requirements.

I think this statement is extremely poignant, "Because you're looking for a sense of security when you're spending your budget on a web app". I think a big problem is lack of knowledge on the part of the client. Most people just have no idea what it takes to build web apps. $1000 (or even $5000) just doesn't go very far. But without learning and experience on the customer's side, how are they able to bound the project to understand the potential costs?

I should maybe have indicated that the clients we're talking to usually have between $20,000 and $70,000 to spend. They employ people who're expected to be experienced.

This issue is why I'm not a contractor. I can't deal with estimates because they are always wrong. People claim to understand that, then fly off the handle when it comes up wrong.

Overestimating on purpose isn't the right way either, and not just for ethical reasons. Finishing early makes the client think you cheated them, and actually working the whole time makes you think you cheated yourself. Especially if you end up having underestimated after all.

And changes? Ugh. Clients truly don't understand why X is harder than Y, and they think it should cost the same.

I've been pretty successful in the past at guiding management into solutions that are easy to code and maintain, rather than the nightmare they asked for, but having to explain all that to a new client each time would be too much.

I completely agree. It's tiresome to repeat yourself over and over, specially when you know it's also in the clients' best interest to not work with someone who guesses (and therefore often overestimates to be on the safe side). I'm hoping our prospective clients will read this blog post first and then -still- want to contact us. They are the gems you want to work with.

To be fair there definitely are people who get burned several times by eLancers and cut-rate web shops in their town and eventually realize that they get what they pay for.

Personally I wouldn't do flat fee work unless the profit margin was astronomical or there was some other factor that made it heads-I-win, tails-I-win situation (this doesn't mean the client has to lose.)

Some kinds of work are eminently estimatable, if the requirements are stable. One time I was working at a job shop for a client who needed a PHP CRUD app. Because this was built within a framework, I was able to break the requirements down and estimate tasks to within 15 minutes and be able to schedule a week's work with about 10% accuracy.

The one trouble is I'd call up the client and show them what I was doing and she'd talk for 2 hours on the phone (which we'd bill $240 for at our rate) and have an endless list of small and medium-sized tweaks she wanted. Had I not put my foot down, she would have blown the budget and schedule completely. I did put my foot down and she fired our agency.

Fixed fee projects create moral hazard for the client: if you're really good in your sales process you might be able to get 4 out of 5 clients to be reasonable and make a profit on them. If you want to keep the 5th client you'll have them overrun the budget by a large enough factor that you lose what you made on the other 4.

An estimate should always be a range. And it should always cover the "cone of uncertainty".

If you don't know, reflect it in your estimate. Get the client prepared for falling within the range. And comfortable with the cost on the far end. Because while you hope it doesn't happen, it very well could.

In my experience a range of, say, £5000-7000 usually means £5000 to the client and it can take hours/days/weeks to budge that perception no matter how initially prepared they seem. I've also found people who, while comfortable with the upper estimate if it goes there, will then expect freebies on top as recompense for making them pay 'extra'.

Contracts and making sure you have fully prepared the client go a long way here.

If they consistently believe/say/whatever the low end, you have not done your job in preparing them. Every single time they mention the low number, you mention the high number.

If they expect freebies or feel as if they are paying 'extra', you have failed in setting expectations.

Amazing how often this happens to even large corporations. Sign a very expensive contract with a supplier at a "flat fixed fee", requirements are never detailed enough and always changing. Much better off paying time and materials or create a partnership where both benefit from the outcome

Reply on news.ycombinator.com