Web Standards ROI
June 01, 2004 | 
35 Comments
Is there anyone out there who still doubts the real world benefits of developing with Web standards?
If so I’ve got some pretty compelling evidence for you. Especially when it comes to the benefits of CSS based design. I’m going to show you some actual numbers that show pretty clearly the Return On Investment (ROI) of going with Web standards.
What I’ve done is gone back through my invoices and pulled the numbers for two very similar projects (in terms of scope and deliverable), one done back in 2001 and another done just before the end of 2003.
I’m going to show you the differences between these projects in terms of time, effort and dollars as well as talk a little bit about the real world effects of using Web standards. It’s really pretty amazing.
As an extra I’ll also talk about a project we recently completed for the hospital. It’s not as striking numbers-wise, but it will still show a substantial benefit of going with a Web standard development process.
Breaking Down The Projects
What I’m going to do is show you some numbers that will give you an overview of how the project went. Most of these are accurate and I’ll note where I’ve had to estimate. I’m also just going to show relavant numbers, not every last thing for each project. For illustrative purposes I’ve placed my hourly rate at $100 an hour. The actual rate for these projects was different for each of them, so that will be the baseline for comparison.
After I show the numbers I’ll talk a bit about the project and give some insight as to what those numbers mean.
All of these projects were simple brochure-type sites. I did use server side includes on both of them, but other than that they were, for the most part, static and didn’t need much in the way of custom or complicated development.
Project #1 — 2001
The Numbers
Pages: 56
My Total Hours: 112
Hours spent in Web development: 80 (initial build was around 40 hours and there was about 40 hours of tweaking.)
Total days spent working on it: 195
My cost (@ $100 per hour = $11,200 total. $8,000 for Web development.)
The Details
Obviously this project took quite awhile. 112 hours total for 56 pages! It also took about six months from start to completion. Sure, I actually made good money on this, but when you think about how long it took, it doesn’t sound like a very good deal any more.
It was built in a rather “old school” manner with tables, images and very little CSS. As tends to happen with these projects my client changed their mind numerous times.
The IA shifted, labels changed and content was reworked. In general this client was no better or worse than any other, but because of the labor intensiveness of non-standard development and the fact that I hadn’t used CSS, I spent quite a bit of time in Photoshop reworking navigation images.
The tabled layout, as I remember it, was fairly easy to implement at its inception, but grew harder and harder to tackle as time went on.
This project was fairly typical for projects I used to work on before I began to develop with Web standards. How times have changed — for the better.
Project #2 — 2003
The Numbers
Pages: 82
My total hours: 32
Hours spent in Web development: 19 (Initial build was under 10 hours and there was about 9 hours of tweak. This doesn’t count content build which was about 8 hours.)
Total days spent working on it: 15
My cost (@ $100 per hour = $3,200 total, $1,900 for Web development.)
The Details
This project was very similar to project #1 in many ways. It was a mostly static site that needed quite a bit of tweaking. To be honest, the client here was a bit more organized and did respond to my e-mails quicker, but in general they weren’t much different that the client from project #1.
As you can see there was a pretty substantial difference as far as time, effort and money goes. This project cost about $6,000 less to build, took 32 hours total and was completed in less than a month. This project was also much larger in terms of content and pages.
Once I had the template built it was very easy to make changes across the site, so I could concentrate on building the content out and let the UI and IA tweaks come in any old time. I can’t emphasize how much of a difference this made.
Now obviously I made much less money on this one, however, if you look at the time spent it works out better. If I could do 12 projects like this in the same time span as project #1 you can begin to see how that benefits me. As well, since I’m passing a substantial saving off to my client I could justify a rate increase.
Return On Investment
I think it’s pretty clear that developing with Web standards can save time and money. I will admit that there is a bit of a learning curve, especially when it comes to CSS and there are also times when trying to support IE’s botched handing of CSS can add time to a project, but even so — it’s well worth it.
To further illustrate this point I’ll talk about a project we did down at the hospital. The benefits aren’t as tangible here as this project was worked on by a multitude of people and we had quite a bit of legacy content to deal with (made the content build out much harder) but I can give you a general idea of the numbers here to give you another case to compare with the first two.
Project #3 — 2004
The Numbers
Pages: 65
My total hours: 65
Hours spent in Web development: 15 (Initial build was under 8 hours and there was 7 hours of tweak. Content build was long and hard — about 30 hours*)
Total days spent working on it: 32
My cost (@ $100 per hour = $6,500 total, $1,500 for Web development.)
*much of this was due to legacy code and revisions
The Details
This project could have taken much longer than it did. Because our internal Web team was caught in the middle of an internal client and an external vendor, we could have easily been crunched between the details of the design and the constant tweaking of IA and content.
Because I built the site with CSS in mind I was able to begin on the content and built out of pages before the design was fully completed. This allowed me to get moving on the site.
I spent much of my time on this project on things other that Web development. Thus the difference between my time and the Web development time. If I’d had to spend lots more time developing this project wouldn’t have been smooth at all.
Ideally there would have been more interaction between client and vendor, but as it was our Web team had to make the best of an awkward situation and keep this project moving. CSS and a Web standards based development allowed us to do this.
At the end the biggest obstacle ended up being the conversion of legacy content, and to be honest, we could have gotten around that had we needed to. Luckily we had plenty of time and actually made our deadline with relative ease.
True, it wasn’t perfect, but had I had to wait for the design to be finalized or not had the power of CSS to help mitigate the changes — both with design and with content — this would have taken much longer, thus costing the hospital more money on several fronts.
All in all this took just over a month to do. That is amazing when compared to other projects I’ve worked on at the hospital. Web standards help facilitate that and in that way helped to effect a positive change in our process.
The Bottom Line
It’s can be pretty simple when you break it down — web standards can save you and your clients time and money. It’s not a silver bullet, but it’s a huge step in the right direction.
Filed under: Web Development
Comments
1. JC said:
Very useful information, Keith. Thanks for posting it!
Posted on June 1, 2004 01:18 PM | #
2. Bryan said:
I agree. Great info Keith. I can see this making the rounds among bloggers everywhere. This is what we have been needing. NUMBERS! Clients want Numbers, development teams want NUMBERS! Not jargon. There has been too much JARGON with web standards and its benefits and not enough REAL WORLD benefits to back it up.
Posted on June 1, 2004 01:28 PM | #
3. David Barrett said:
And they use CAPITAL LETTERS in their copy to SHOW EMPHASIS. :)
The ability to work on the content, in (X)HTML form, before the design is complete is a real advantage of CSS. It’s tarnished somewhat by the extra markup required by some techniques (an extra div here, and extra div there), but it’s nowhere near as bad as hacking through nested tables.
It also means that new design ideas can occur to you while working with the content that you can (usually) implement without serious hassle. Think that colour you chose for the sidebar could be toned down a bit? Ha! Now edit 50 bgcolor values on those td’s!
Posted on June 1, 2004 02:00 PM | #
4. daniel said:
I’ve noticed this as well. I mean to say that using a css based design in my little projects allows me to edit pages easier. Whereas when I used to use tables alot, some pages would not display correctly due to some stupid little mistake. With stylesheets, I’ve been able to have a uniform, almost bug free design that is pretty easy to manage. Thanks for the story. It really pushes the thought in my mind even more to use css and standards.
Posted on June 1, 2004 02:14 PM | #
5. Sérgio N. said:
Have you already written anything about writing proposals and figuring out quote numbers for clients?
I liked your text.
Posted on June 1, 2004 03:07 PM | #
6. Phoat said:
Very good article. Being new to the whole web developement thing, I had trouble figuring out what to charge my clients. I wonder tho, how to you estimate how long it will take you to complete a project? Doesn’t the client want to know a ballpark figure before you tackle a job?
Like Sergio said, I’d like to see an article on the documents involved in a project; from initial proposal, to estimates, to sign offs. Have anything to share about this? I know it would help me out enormously.
Posted on June 1, 2004 03:27 PM | #
7. Keith said:
Sergio and Phoat – It’s a bit off topic for these figures, but I don’t think I’ve written much about my process with estimates and proposals and such. I do have one, and it’s worked well for me, but to be honest I’m not sure I know the best way to do this.
I will say that every client is different and while I do have a fairly set hourly rate, it can be challenging to accurately estimate projects.
For a few things you can try gotomedia’s resource pages. There are some good resources there, particularly the client survey and creative brief stuff.
Maybe I’ll talk more about this in the future, but it should be in a different post to keep things relevant.
Posted on June 1, 2004 03:36 PM | #
8. Ian Firth said:
How exactly do you bill your time ?
You state some very bizarre numbers there, 32 hours over 15 days, 112 hours over 195 days, 65 hours over 32 days… Do you only work a half hour a day, or do you have multiple projects running concurrently? :)
Did you have existing designs to start with, or was it all from scratch? I’ve spent 40+ hours just designing the look/color/feel of a site and can’t imagine building an 82 page site in just 32 hours (3 days for me).
I don’t design completely to web standards, but once I have a design completed, it’s effortless to turn that layout into a site of any scale, but not in the (billed) timeframe you mention.
Posted on June 1, 2004 09:51 PM | #
9. Micha Schopman said:
I am always missing relevant information which this type of posts. What was the design, what was it’s complexity and such.
Everybody can claim high returns on investments, but it is like comparing oranges and apples, it all depends on the design. Creating a tabeless google.com or creating a tabeless msn.com, a big difference.
Posted on June 1, 2004 11:20 PM | #
10. Philipp Bunge said:
I agree Micha.
Furthermore, I believe that it is impossible to make any general assumptions on the profitability of using web standards by comparing any two similar projects.
For a more accurate analysis, you would have to compare numerous projects of various complexities. (Then again, how do you estimate complexity?)
Another question is also the timespan under which to estimate a TCO for the customer. Considering that for content managed sites designs become outdated before content does and most sites being developed are content managed (?), applying web standards might be more profitable customers in the long run. Then again, for static pages this might not be so.
Nonetheless, thank you for your article Keith. I think there is a need for more analysis of the advantages of web standards for customers.
Posted on June 2, 2004 12:17 AM | #
11. Pete F. said:
Thanks for the insight, Keith. It’s very helpful to see a Standards advocate to print actual figures, both in terms of cost and development time—there’s little else that stands out better when dealing with speculation and doubt about employing new processes. Much appreciated.
Posted on June 2, 2004 12:37 AM | #
12. Joshua Porter said:
Hey Keith, you make a strong case for designing with web standards. Given numbers like yours as well as experiences like ESPN’s, I can’t help but think that most sites would improve by developing this way.
That’s why I’m always a bit confused when large, templated sites like Amazon and Ebay don’t create new pages or templates this way. If they were convinced of a decent ROI, then one would think that they would have made the change by now.
Giving them the benefit of the doubt, perhaps they are aware of factors that reduce the return?
Posted on June 2, 2004 07:42 AM | #
13. Randy Charles Morin said:
Examples are evidence, not proof. But I agree w/ the premise.
Posted on June 2, 2004 08:25 AM | #
14. Rune said:
Great article!
It sounds very logical at all. You must get cash :-)
Posted on June 2, 2004 08:56 AM | #
15. Keith said:
Ian – I do have multiple projects at the same time and yes sometimes I only work a half an hour a day. I freelance in my spare time, I have a full time job. I also work pretty fast.
All of these projects were built from scratch. I designed the first two myself and the last was was designed by someone else, I just did the build out. I’m really trying to show Web development numbers here, not design. Web standards (mainly CSS) greatly decrease development time once you’ve gotten past the learning curve.
Micha and Phillip – Your questions and concerns are valid. I tried to explain some of what you are asking about in the post. True it’s hard to gauge the actual ROI without understanding all the details.
The first two projects were very similar in scope, design and complexity. That is why I chose them. Regardless of the details the difference number-wise is so huge you can’t help but see some ROI. For the rest, you’ll just have to trust me when I say that a Web standards development process is what made most of the difference.
Posted on June 2, 2004 09:30 AM | #
16. Keith said:
I should also point out that there can be quite a few factors as to how long (in days, not hours) these projects took. These numbers are real, that’s why they don’t always make sense time-wise. That first project took six months because of lengthly development time and intermittent communication with my client.
That is the reason why I put the third example in there. A project like that could have easily taken months to complete because of the various stakeholders. CSS gave us the ability to work on the project and keep it moving forward instead of having to wait for every decision to be finalized. This was huge.
Posted on June 2, 2004 09:59 AM | #
17. Graphiz said:
I totally agree with you, that ROI on CSS based project is more. Also the possiblities of CSS designs are infinte.
Good post.
Posted on June 2, 2004 11:17 AM | #
18. Joseph said:
Keith — Thank you for posting this information. I have found it most helpful and illuminating. The time you show on web development for the CSS-based projects is amazing to me. The numbers are so low in terms of hours! I have a long way to go to speed my process up. As always, keep up the great content.
Posted on June 2, 2004 11:28 AM | #
19. Mike Wilson said:
Hi,
I’m still not convinced. The only way I can see this type of numbers comparison being realistic is to develop a single test site using both methods. Even then, using a different developer for each method will skew the results to some degree, so it would have to be done by one person or one team. The two year gap in between the projects also adds to the unreliability–you’ve likely learned allot over those two years and that has certainly improved your performance. Even if we are just talking project planning techniques, and not the actual coding, the added experience will skew the results in favor of the more recent project. Officially, I am in the standards corner and have been trying to move my company towards this for some time, but honestly I cannot tell my CEO that initial design time (while achieving our level of design goals) will be reduced through standards based development. I prefer to speak to him on issues of reaching every consumer we can with one effort–this is where standards does pay off. Maintenance is certainly easier, but I digress.
Posted on June 2, 2004 12:10 PM | #
20. Keith said:
Mike – Fair enough. But these numbers are reliable in my mind. Then again, I was there.
Also, for the record and to shed a little light on the gap between projects. I’ve been design, developing and building Web sites for about 10 years now – so it wasn’t like I was a newbie back in 2001. There was an increase in knowledge and that increase was Web standards and CSS.
As to your point about talking to your CEO about other benefits, that is a great idea, just don’t think for a second that I feel these numbers reflect the only benefits of Web standards. I just wanted to show some real numbers to back up what I think is a huge ROI benefit on several levels.
I also do plan to post some ROI numbers on maintenance as soon as I have time.
Web standards and CSS mean that I have more time to work on more projects. Period. Milage may vary, but in my experience this is the case.
Posted on June 2, 2004 01:12 PM | #
21. Mike Kear said:
Mike Wilson, you’re dead right. This is a whole lot of black magic nonesense. You stick to that view and pick holes in the detail. Suits me fine.
Because all the while you do that, I have an enormous advantage over you, and you are missing out on a change that can revolutionise the way you build sites. I switched the development methods in my shop to using web standards in January, and while I’m not going to get into debates about the nitty-gritty of hours spent etc, but I can verify that my experience is even more dramatic than Keith’s. Using web standards allows me to separate design and development and content. As Keith says, I can be building while design is still proceeding. The client can change his mind dramatically and I dont have to panic, I can accommodate the design changes quite easily for the most part. Certainly with changes to only a couple of files, not like changing scores of files which is what would have been required last year.
But most importantly, I run a small shop - only two people here - and I can take on far more work than I’d have even contemplated last year. Using web standards has made a direct difference to my income.
Furthermore, I can feel confident I’ll be able to support the sites easier in the future as the client’s needs change because the whole structure of the site is far simpler. I’m not talking a bit simpler. I mean FAR simpler.
You keep nit-picking and denying the advantages along with the others who have their heads in the sand, and I’ll be busy making more money while you do.
Keith I dont know whether my numbers would be the same as yours, and you’ve given me some food for thought in that, but I can testify from my own experience that if you’re wrong in your numbers, its through UNDERSTATING the difference, not overstating it.
Posted on June 2, 2004 03:08 PM | #
22. Thibaut Allender said:
Sure, CSSize or die ;)
It only took me a few hours to design a second skin for my website, and the client just have to download the new css and some background images… that’s it!
Posted on June 2, 2004 04:21 PM | #
23. Mike Wilson said:
Hi,
Wow, thanks for the great flame Mike Kear. :) In all fairness, I never said I didn’t use standards based coding (I advocated it in my comments), I just don’t feel those types of numbers are entirely realistic for the reasons I previously stated. I am sure Keith is being completely honest from his experience, as I am also sure mileage will vary depending on the scope of the projects involved. In some situations, it is possible to save money on initial design, but we are talking about 80 full hours here… $8,000. I’ve used both methods (and mixed) on similar projects as well, but I cannot report this kind of savings–money or time. I would like to see a test involving parallel development methods within a single project–same goals, details, changes, and project specs. That would certainly put the point to rest. I don’t have my head in the sand Mike, I was coding to standards when WASP was first formed “way back” in the HWG glory days and being flamed for advocating it. I use a very thorough development methodology that captures many of the problems you feel are important, before I ever write a line of code–very rarely do my clients backpedal on me after I start coding. As far as I am concerned, software or application coding is the least likely thing to hold up any project I’m working on. I do understand your passion and your point of view and I completely agree that standards based development is the best route for most projects and the future of Web development. Finally, as I said before, maintenance is an entirely different issue and I agree completely that standards based coding is far more cooperative when it comes to updates and additions.
Posted on June 2, 2004 07:21 PM | #
24. Ben Wong said:
Nice article, Keith.
I’m a fan of standards; I’m the self-appointed CSS guru at work and I cringe when I see <FONT>. However, I’m going to have to agree with Mike W. You’ve probably pulled a few people off the fence, but I think you’ll need more case studies and more detail to make people jump over it. The learning curve you mentioned has a cost too. Seems to me like you compared the “operating costs”, but neglected to mention the “capital investment” needed to make the transition.
Posted on June 3, 2004 05:32 AM | #
25. veen said:
Excellent justifaction for using XHTML+CSS on fixed bid projects, rather than time and materials. Or increasing your hourly…
Web standards = margin!
Posted on June 3, 2004 09:23 AM | #
26. scott hirsch said:
Wow! Interesting and insightful analysis.
One point strikes me. You’re only talking about a cost analysis for particular projects at time of build – there are much more savings to be realized in future tweaks, adding pages, minor redesigns, etc.
That’s potentially a huge component of “return,” and I’d love to hear more about how you would propose estimating it.
Posted on June 3, 2004 10:46 AM | #
27. Mike said:
*Bookmarks this page for future reference*
I love it! I feel the urge to give a presentation here at work now (Northrop Grumman) talking about the benefits of web standards and the like.
Like they really care what I think though :)
Posted on June 3, 2004 12:41 PM | #
28. Brad Daily said:
Mike from your comment a little earlier, you might find this interesting. Sounds like this is what you are looking for…
Tables vs. CSS - Fight to the Death
Fictional project, but a good read anyway…
Posted on June 3, 2004 04:55 PM | #
29. Mike Wilson said:
Hi,
Brad - thanks for the link. I did read the article, but wasn’t as impressed as I’d had hoped I would be. Standards won, which is good as that was the goal of the article, but the writing was biased from the opening page, so I didn’t expect otherwise. I don’t really equate non-standards design with the absence of CSS. I mean, using no CSS is certainly against the grain of the standards movement, but in all actuality, you can design a standards based (although not very pretty) page without CSS just as you can use CSS and still be non-standard. Regardless, his code in the table version is not very well thought out-way more nesting than necessary. His stated time is ridiculous as well. Being generous, I feel I could have finished up that layout in about an hour–two hours if I took a break or a nap. Don’t get me wrong though, I was rooting for the CSS version. :) It is much cleaner, a bit faster, and will certainly be easier to maintain. I am not anti-standards, quite the opposite, I just don’t agree that faster initial development is one of it’s stronger points. Achieving a well thought out, semantic, and attractive page takes time, but I do agree that the ROI will come down the road.
Posted on June 3, 2004 09:40 PM | #
30. Eric G said:
Good posts. However, every client and project is different. After you’ve worked on plenty like some of the guys here, they’re bound to find similarities.
You can have a client that doesn’t give a hill of beans about CSS/XHTML, they want their logo yellow! Its about managed expectations… what’s important to the client – yes easier maintenance, SEO, etc. But at the end of the day they need to be comfortable with YOU the designer/salesperson and the PERCEIVED quality of your output (if its in XHTML/CSS, bonus). That’s what counts.
I’ve seen too many guys bang their heads against the wall trying to sell this stuff to CEO/CIO/CFO types only to get frustrated (especially in-house). You have to juggle both sides of the fence. Match their priorites and expectations even if you have to “standardize” the code under the radar and behind the scenes.
You can type using 2 hands instead of picking with one finger? Your code actually validates? Good for you! Now make sure your copy is airtight, the branding consistent and help shuttle the project thru the compliance department. Now you’ve graduated from ROI to Value-Added!
Posted on June 11, 2004 10:40 PM | #
31. Che Tamahori said:
I think I’m in Mike Wilson’s camp on this one.
I also work at a company where we’ve got a process that tries to remove any ambiguity from the IA long before we open up Photoshop (let alone start coding a page). So the ability for us to tweak things as we go is less critical.
Additionally, we often work on CMS-based sites, which greatly reduces the flexibility and content/presentation seperation benefits of CSS. You want to make radical changes to the site design? Change a couple of templates and re-render the pages.
We have a commitment to accessibility and cross-browser viewability. We find 100% pure CSS sites to require _more_ testing – not less, if they are to degrade reliably through older browsers.
And this sounds like sacrelige, but table-based layouts are very reliable for creating page structure!
When people start talking about FONT tags, you know were talking about a straw man situation. Only people with deep psychosis use FONT tags these days. The article Brad points to makes a point of highlighting the evil horrors of a no-CSS approach.
But does nayone seriously eschew CSS these days?The real debate is whether you:
1) use CSS alone for all layout, or
2) still use a handful of TABLEs tags to implement some basic page structure, and CSS for everything else (typography, borders, backgrounds, etc)
The first approach leads to the inevitable degradation problems with floats, box-model hacks, and the rest. Which you work around. Which is okay.
But what is the problem with the second approach? Seriously? Especially when you’re using a templating system?
I’m another of those people curious about why people like Google and Amazon continue to use table-based layout. Could it be that it continues to work in all browsers (including the new ones), whereas CSS only works on the new browsers (and those poor souls with pre-v5 browsers have to eat cake)?
Posted on July 18, 2004 10:38 PM | #
32. Nic said:
Hello, I just wanted so say thank you guys ! I really like your site and i hope you’ll continue improving it. Kind regards by Nic
Posted on July 21, 2004 06:23 PM | #
33. Leslie Elledge said:
URLs for these mythical clients’ sites? I am a designer (as well as a technician) and there are still things you can’t do without a table. I’d like to see your resulting sites. Perhaps the actual end-product is VISUALLY not as successful w/only CSS.
Posted on July 25, 2004 10:50 AM | #
34. Will said:
validates? Good for you! Now make sure your copy is airtight, the branding consistent and help shuttle the project thru the compliance department.
Posted on December 26, 2005 08:44 AM | #
35. David said:
I use a very thorough development methodology that captures many of the problems you feel are important, before I ever write a line of code⦣x20AC;“very rarely do my clients backpedal on me after I start coding.
Posted on January 4, 2006 12:20 AM | #
Comments are now closed