Why the calculator?

I have been writing, migrating & QAing web content for more years than I care to remember - usually as part of big CMS implementations.

At the beginning of such activity a project manager will often put 1 of 2 questions to me...

1. "Shane! We need to write 100 pages of brand new content & migrate 500 existing pages into the CMS. How long will that take?
2. "Shane! We have an absolute maximum of 20 days to produce & migrate all content. How many pages can we handle in that time?"
   

Well, it depends...

Neither of those is so easy to answer. There are a lot of dependencies. But after a long time of throwing guesstimes around, I decided to get a bit scientific.

I began to track the time needed to complete the 3 main steps of content production and came up with the following per-page averages:

Bear in mind, these are averages. Some pages take a lot more time and some take a lot less time. But on average, those are the times I now go with.

Of course, those estimates can be lowered if content is not 'completely new' but is being repurposed from some other source, e.g. brochure content being republished online.

No way!

The most surprising and least accepted figure is the 5 hours for writing a page of new content.

Pretty much every client I have ever worked rejects that figure out of hand. They're usually thinking along the lines of 2 hours per page. Yet in reality, it often takes 6 hours or maybe even 7 hours for really good quality output.

In fairness to them, I guess we all had to learn the hard way that great content is incredibly labour intensive.

Into excel

A long time ago I put all these numbers into an Excel spreadsheet so I could work out calculations on the fly - and also demonstrate to clients what they really needed to invest in when planning content. (After all, good Web Governance is all about understanding the scale of an intended site.)



An ex-colleague of mine and I also tried to webify it a few years back. We made a bit of progress and did some nice mock-ups - but then other projects took over, he left and it got lost. Oh well.

No matter. In many ways I actually think it works better in Excel as it can be so easily 'brutalised' into whatever shape you want. That's flexibility is handy.

I admit it is far from perfect and you probably disagree with its estimates. But sometimes all you really need are approximations for planning project activity - and that's what it delivers.

How does it work?

It is pretty self explanatory. On the first tab you can do 2 calculations...

1. Enter the number of pages you need to produce. The spreadsheet will then calculate how long they will take to complete in man-days.
2. Alternatively, enter the number of man-days you have available. Excel will then tell you how many pages you can produce in that time.

(The second tab is the 'backend', where you can change the underlying phase estimates.)

Did you know .... as well as being a full-time shoulder-to-cry-on for governance matters, I am also a part-time dabbler in web development?

No?

Well, over the past few years myself & my brother (Ronan) launched two social-networky-type sites. The first is a community rating website called LikePlace and the second is Wandermates, an app for connecting solo travelers.

Unlike this brochureware site where QA issues are limited to a few broken links or typos, both LikePlace & Wandermates collect oodles of membership & other data from visitors.

Inevitably it happens that from time-to-time somebody will subscribe more than once, or encounter an issue that adds a duplicate to our database. As a consequence I spend a fair amount of time QAing data, as well as normal things.

Thus far both sites are of a small enough scale that automation is not really necessary. I can remove duplicates manually.

But not everyone has that luxury

I know of many charities and retailers that collect the details of hundreds-of-thousands of donors or loyalty-card members into massive databases, including names, emails, postal addresses, etc.

It is also well known that such databases are ridden with duplicates. But they are so huge that no amount of manual effort alone can possibly clean them up.

Something else is needed

Enter Dedupeyourdata, a new QA tool that can dedupe excel & marketing lists. All you need to do is upload your data in Excel, CSV or any number of other formats and let the deduplication software do the rest. Among its features it can:

• Find duplicates in Excel lists 
• Match one list against another 
• Match using sounds-like 
• View all duplicates and scroll through them 
• Remove/delete duplicates and download clean lists 

Just think of the time and effort that can be saved in cleansing data before major online or offline marketing campaigns at Christmas, Hallowe'en or Thanksgiving.

As an aid to the maintenance of clean data I consider it an extremely useful addition to the suite of Web QA tools.

I first became aware of Siteimprove back in 2004. I had just been appointed Webmaster for a leading website and needed a bit of help to get Quality Assurance under control. Enter the chaps from Copenhagen who had exactly the tool I needed.


After that I always kept an interest in how things were going at Siteimprove. As such, I was delighted a few weeks ago to catch-up with Torben Rytt, CEO of Siteimprove's US operation.

With so many clients at the coalface of online, Torben's insight into current governance pressures & trends is second to none.

But putting techie issues aside, I began our chat by asking how a Danish guy like him could end up in the Mid-West of America..?

Read the latest Web Governance vendor interview with Siteimprove.