Using the 3×3 rule to redesign our intranet

ne of the greatest frustrations of an intranet that organises its content to reflect the organisation chart is that it requires a great deal of prior knowledge in order to find what you want. You need to know, for example, that Organisational Development is responsible for senior executive development, rather than, say, HR or Training.

When we came to redevelop our part of the intranet, we knew we needed to rethink our information architecture from the point of view of our users. This is always a challenge as an insider: prior and in-depth knowledge can be a liability.

Every book and how-to guide to information architecture will tell you to begin by identifying your users and determining what they need. This is always good advice. The challenge comes in determining who your intranet site is for. It’s easy to say “employees of the organisation” but that’s circular since, after all, only employees of the organisation have access to the intranet.

The question is, should our intranet prioritise support for our staff (i.e. those who work in our area), or prioritise its value as a”promotional site” to sell our services, and our value, to “visitors” i.e. the rest of the organisation? The answer, unsurprisingly, is “yes”, although in the end discussions revealed a preference for staff over visitors.

These two goals seemed at first to be at odds. To be useful to current staff the architecture needed to enable access quickly and accurately to the tools, products and information they need daily. As a “promotional” tool for non-specialists, it needed to showcase the range of services we provide, provide information on how to access them, and provide guidance on how to integrate what they do with what we do, all without burying visitors in detail they don’t need or want.

My approach to reconciling these two goals came out of a rule I had come across several years ago for presentations. Called the 3×3 rule, it advocates making three main points:

1. Point 1
2. Point 2
3. Point 3

… and if you have time, enlarge upon each with three points about each:

1. Point 1
   a)
   b)
   c)
2. Point 2
   a) etc.

The 3×3 rule means that if you have 5 minutes to make your presentation, you can communicate your main points; if you have 15 minutes, you can illustrate each with three sub-points; and if you have 30 minutes you can develop the subpoints in more detail.

Applied to the intranet, this meant that the first level of navigation needed to communicate the main points, points that both visitors and staff could grasp immediately about how we work and what we do. This top level needed to communicate the scope and breadth of what we do without any detail, but with enough coherence between the labels to outline the structure.

Second level navigation labels would flesh out the structure, and content would fill in the detail.

Nav labels obviously are extremely important in ensuring this works. Although in general jargon is a very bad choice for nav labels, it can serve a useful purpose in teaching visitors about the language of the specialism. This only works if the jargon is not too far removed from common language usage.

For example, we call the documents and reports that we produce “products”, and ones that are signed off are “released products”. It’s not too much of a stretch to expect a visitor to understand these terms when they visit our site. Why not just call them documents? Because that’s not what we call them. Visitors might recognise the term documents more readily, but when they talk to our staff, they will hear about products. The labels are a quiet way to educate our visitors, and inform them about what is unique about our business.

In a large corporate site, the 3×3 rule couldn’t be applied literally without generating endless levels of navigation to dig through, but the principal remains a useful one.

Getting it wrong after it goes wrong

I got prompt responses from Fresh to my enquiries about payment and fixing my duplicate order. Unfortunately, I got the same response to both problems, which only reduced my already shaky confidence in the process. Worse, neither fixed my problem in a way that worked for me.

I was told that the NZ site doesn’t accept credit card payment, only PayPal or bank transfer because “PayPal is the safest way to pay”.  They suggested that if I wanted to pay by credit card I could use the Australian site which does accept credit card payments.

This of course makes no sense at all. My first thought was, is there some reason why Australians are happy with “less safe” ways to pay, but NZers aren’t? And if credit card payments are so risky, why do so many huge e-commerce sites (Amazon for example) use them?

My enquiry really boiled down to “why am I being inconvenienced?” Their response attempts to reassure me that the inconvenience is for my own good. Their alternative, should I want to use a credit card, would inconvenience me differently as I would incur exchange rate costs and much higher shipping costs.  None of which would, in fact, make it any less inconvenient for me.

But inconvenience is only one part of what’s gone wrong here. There are fundamental problems with their payment process:

1. The payment system didn’t work for me at all – I got an error message and this wasn’t acknowledged in their response at all
2. I was led to believe I could pay by credit card because the brand images for Visa and Mastercard were there beside the PayPal image. There was nothing to tell me these options were not available from the NZ site
3. There was no warning that I was irrevocably committing to a transaction by clicking the payment button.

The first issue is a serious one for them, but the last issue is potentially the more serious one for customers. Apparently there is no cancellation possible once an order has been lodged, a fact that needs to be made apparent at a very early stage of the purchasing process. If there is no ‘bail out’ option, customers need to be warned of this before they begin.

I had assumed that clicking the payment button had lodged the order, even though the payment did not go through. However, I was alarmed when I realised that orders appear to be considered as irrevocably lodged earlier than this, at a stage in the process I was unaware of: the first order I assembled but never took to the payment stage was also recorded as “ordered”.

Many e-commerce sites handle the transaction process extremely well, with clear advice to the purchaser along the way about when the order is ‘committed’, what stage of the ordering process they are at, and almost all allow or prompt for some form of confirmation immediately prior to the point of no return. All of this clarity reassures the customers that they are in control of the purchasing decision. It mirrors the physical shopping experience: you know when you’ve paid for your purchase, you know if returns or exchanges are allowed, and you know can change your mind any time up until you sign the credit slip or enter your PIN number. You know when you’ve reached the point of no return.

I won’t attempt to purchase from Fresh again. Their selection is good, their prices are good, and I understand their delivery is fast, but their transaction process is poor enough and ambiguous enough to leave me feeling that it’s too risky.

Getting it wrong: new visitor experience

I’m fairly loyal when it comes to on-line shopping sites. I usually research a site fairly thoroughly for its selection, shipping costs, and privacy, returns and cancellations policies. So when I decide to register it’s usually the start of a long-term relationship.

But sometimes, the first date is almost bad enough to drive me away forever.

Today I had a first date with Fresh. Here’s what I did. Fairly typical first date behaviour:

• I got part way through the checkout process then realised I should have signed up for the loyalty programme
• I went back to sign up, signed in as instructed, then returned to my shopping cart
• I clicked the Pay Now button to finish my order.

Then it all went wrong. I got sent to a PayPal-hosted error page telling me I couldn’t complete the transaction. The payment instructions said I’d be sent to a third party site where I’d have the choice of credit card or PayPal, but all I got was PayPal, and a PayPal that wasn’t being welcoming. Besides, I wanted to pay using credit card.

I tried again, and again, but no, three tries later and still the same result.

So I found the customer support form and sent a message saying what had happened.

Then I returned my account page and discovered TWO orders were showing, one from before I joined the loyalty programme (i.e. without the loyalty club discount) and one after. Both orders said they were in processing, even though I hadn’t paid for either. Worse, I had never even clicked Pay Now for the pre-loyalty order.

Back to customer service to send another message.

I’m now waiting to find out what will happen. What they do next will make or break this relationship.

I had a rocky first date with Fishpond, but all was forgiven and we’ve gone on to have a fulfilling and very happy relationship. We’d had a hiccup on my first purchase, but they fixed the problem and gave me a voucher. I’ve loved them ever since, and all for the cost (to them) of a $5 voucher.

Three things here:

1. First impressions count. Giving me a good experience browsing your website is all well and good, and flash (literally and figuratively) graphics may be seductive, but if the purchasing process isn’t foolproof, idiot proof, change-your-mind proof, then it’s all a waste of effort. If the one thing your website is supposed to do is sell stuff, then that is the one part of the site that HAS to work. No exceptions.
2. If, however, something goes wrong, and even if that something is because the customer is an idiot, what you do to fix it is your second chance. Even if you’ve tested your checkout process a million times and are sure it’s idiot-customer-proof, you still need a backup plan to make sure the affected customer goes away thinking your customer services was amazing.
3. There is no third chance. If you don’t get 1 and 2 right, it’s the end of the relationship.

So Fresh are now up to number 2. What happens next will determine the fate of this relationship.

Challenging “longer is better”

The more long business documents I have to read, the more convinced I become that longer is not better.

The conviction that long means “I did lots of work on this” is deeply entrenched. At school and uni, the first question about any essay assignment is always “how long does it have to be?” and in truth, the answer is always the same: “As long as it needs to be and no longer”.

However, “as long as it needs to be” is usually given an explicit word count. And from that point on, we associate the importance of what we write with length. It’s a paper worth 10%? About 1500 words. A thesis that represents the culmination of 3 years of research? At least 10,000 words, surely…

I am thinking that what we really need with our written products is two versions: the long, fully-documented research report, and the short digest for public consumption. The long report records the research, decision making process, weighing of pros and cons, exploration of alternatives etc, but this document should not be confused with the document that decision makers get. They should get the digest: the short, pithy, to-the-point conclusion. The digest should not exceed 5 pages.

Both documents are necessary but they are for different audiences. Other researchers will want to see the full version, decision makers might occasionally want to see it (perhaps simply to reassure themselves that it exists), but no-one should have to read it.

In theory, everyone should go for this idea. But I am sure they won’t. Decision makers will be nervous about receiving a 5 page report that represents three months worth of research. Can they commit funds to something based on 5 pages of evidence? And the researchers won’t like it. How can they showcase how much work they did in 5 pages? What about that long document they’ve written, what use is that if no-one reads it?

Well, yes, good question indeed. No-one reads it now. It’s just half a tree’s worth of reassurance that someone has Worked Hard. Five pages just doesn’t look like it took enough time to write, even if it took twice as long and is twice as useful.

Don’t make your reader “work it out”

Excellent advice from the BBC (PDF, style guide):

You may ask:“Why bother? The listener or viewer
can easily work out what I mean.”
Listeners and viewers should not have to work
out anything.

Add “reader” to “listener or viewer” and this says it all.

Fun with formulas in Word

Using formulas in Word tables seems to be an unfamiliar feature. Formulas are very handy: they save manual calculations on tables (particularly when you’re updating documents and forget to recalculate totals in a table), having to import calculations from Excel, opening up Calculator… you get the picture.

So where do you find them and how do they work?

In Word 2007, once you’ve created a table and the Table Tools tab is showing on the ribbon, click on Layout. The Formula button is on the far right:

When you click on this, a dialogue box opens. You can paste a function in from the Paste Function option or just type it in. The formulas are limited (nowhere near as sophisticated as Excel) but they are perfectly adequate for adding up columns, even calculating percent of a total e.g. to add tax or factor in a contingency sum. You can also set the format of the result, so if you want dollars select the “$###0.00″ format.

Here is a table with the results of formulas in the Total column:

Word table with formulas

… and here’s the table with the formulas displayed (right-click a formula and select Toggle Field Codes to show this):

Table with formulas displayed

Don’t be put off by the apparent complexity of the formulas. The first part specifies this is a product i.e. it multiplies the columns to the left; the second part (after the “\”) is the formatting instructions.

Something to be aware of: the formula will not keep going past a blank cell. The formula will look in the direction specified (i.e. left, right, above, below) only until it encounters a blank cell or the edge of the table. So if you want blank lines in your table for display reasons, but you need to add up the entire column, enter a 0 in the total column and set the font colour to match the background so it doesn’t show.

You can manually add calculations into the formula. For example, if you wanted to add GST of 12.5% to your total for each row, you can modify your formula to do this:

Formula with multiplier inserted

If instead you want to add 12.5% to the total of the column instead, you will need to do one of two things:

• Use an absolute cell reference to tell Word which cell to multiply by 1.125 – this is fine if you are not going to add or delete rows from your table otherwise you have to remember to modify the cell reference each time you add or delete a row;
• Or, create a bookmark and reference the bookmarked cell in your calculations – this is a better approach if you are likely to be adding or deleting rows in your table and want the formula to keep working without any further attention.

Here is the table with tax of 12.5% calculated on the previous total. In this example I’ve used an absolute cell reference for the total (D5):

Formula to calculate tax on total

Word’s formulas are very useful and fairly straightforward to use. I use them a lot in business cases when I need to show calculations of costs, as they will update automatically whenever I save or print, and that saves me having to laboriously check the numbers every time, reimporting from Excel, managing a linked spreadsheet, or being caught out because I was in a rush and forgot to recalculate something.

Formulas work the same way in Word 2003 and Word 2000, you just find them in a different place: they’re in the Table menu.

Presenting the survey results

I have been wrestling with presenting my survey results. I had to prepare some descriptive results very quickly for a meeting yesterday, but I haven’t gone through all the data yet and only pulled out responses from one section of the survey in order to provide a taste of the results.

I could present some of the data as graphs, and opted for bar graphs like this one:

Bar graph of results

But for much of the other data, graphs weren’t helpful. What I really want is tables. I made up some of my own in MS Word. I did these in Word 2007, not that it makes any difference, but I thought it would be a good opportunity to play with it and see what’s different.

I am fairly sure I could do these in Excel just as easily, although the graphic control in Excel isn’t quite as good and it’s more fiddly to control things like row spacing to allow plenty of white space. Here’s an example of the tables I produced:

Table of results

For a simple table I think this is pretty effective and makes the data easy to see. Looking at several hundred of these will get tiring no doubt but for a first cut this worked.

The initial feedback I got (via a very brief IM with the boss who’s away at the meeting presenting the survey) is that they found the results interesting and want to see more. That’s positive.

Words to watch out for

I take perverse pleasure from designers who describe their work as simplistic.

I am quite sure they have no idea what they are saying. Simplistic doesn’t mean simple, it means reducing the complex to a trite level of simplicity. It’s stupidly simple.

Other words that suffer mis-use include penultimate (this one cropped up in a sitcom I caught the tail end of recently). It means “next to last” and not “best of the best”. Aggravate means “to make worse” although it’s commonly used to mean “to irritate or annoy”. Sometimes it can create what amounts to a Freudian slip: “He aggravates me” might be more revealing than we intended.

The other day a colleague used indictment to praise his team, saying their work was a “strong indictment of how they worked together”. I had to suppress my horror then amusement at what he’d said. Apparently no-one in the room took offence or was confused by this, so perhaps context overrode the word itself. They knew what he meant and so did I.

In speech it seems we are far more forgiving of “flexibility” in word use, possibly because speech moves on rapidly, we don’t dwell on a phrase or a word (except for me dwelling on that “indictment”) and the general context of what is communicated (body language, tone of voice, eye contact) all convey the message, sometimes more strongly than the words themselves.

But in writing, using the right words matters. At the very least you want to be sure that you are saying what you think you are saying.

After the PDF survey, Excel

I have received back all the responses for my interactive PDF survey, and I had a 100% response rate! Okay, fair enough, I only had seven participants.

The Submit button worked as it should, the FDF form data got attached and the emails came in.

I’d noted in the earlier post that processing the form data took a few extra steps. If I had been working with a larger number of responses I would have spent time working out how to automate or at least speed up this part of the process, but because I have only seven responses manual processing was quicker.

Here are the steps I used to get the data into Excel:

1. I began by opening the FDF files in Notepad, then saving them as a .txt file. I tried opening them with Word but it wasn’t having any of it and crashed every time. This may not be the case for later versions of Word (I am using 2002 at work).
2. I opened the text file with Word, then did a Find and Replace to swap the delimiter characters (>>) for a paragraph mark (^p). All of this was necessary to get it into a single column of data rather than a single row. The FDF data is effectively all in one line and when Excel tries to import it, it runs out of room in the first row (I have around 600 possible responses in the survey). If I had been able to import it directly into Excel I would have, then used Copy and Paste Special with Transpose to change the data from row to column.
3. I stripped out one or two lines of irrelevant junk from the beginning and end of the file (extraneous characters that Adobe uses to read the data into the form) and saved the document again as a .txt file (for insurance purposes only, in case I had to re-import it for some reason).
4. Finally, I copied and pasted the plain text into Excel.

Once in Excel I used Text to Columns a couple of times to split the field identifiers from the responses, and then after a bit of miscellaneous tidying up I had my data ready. All up, it took me about an hour to get it as I wanted it.

I’d had trouble with the radio buttons when I was creating the form originally, and I was curious to see what would come back. I am absolutely sure I didn’t do this correctly but I did at least get identifiable results. It seems to have returned only the response to the first radio button. (Fortunately all my radio buttons were pairs! I have no idea what would have happened if I’d had more than two options.) For example, if for question D5 the respondent selected the first option, I would get a result of D51 and if they selected the second I’d see D5. I had to double check results like C11 to make sure this wasn’t a yes to C1 rather than a no to C11. Given the trouble I had with the radio buttons in the first place I was just relieved to be able to tell the results apart at all.

Next step is to make sense of the results.

Declutter your documents Part 3: Content

Decluttering a document in terms of layout and formatting is relatively easy compared with decluttering content.

Perhaps the biggest challenge in decluttering a wordy document is to recognise that it is wordy. Someone wiser than me once said that if you have a phrase in your writing that you are particularly proud of, cut it out. I have always found that to be good advice. The beloved phrase usually amounts to my showing off and contributes little else. So that may be the first place to start.

Do a quick ‘purge’ of:

1. Filler phrases: these almost always have a single word equivalent e.g. “at this point in time” (now), “in the event that” (if)
2. Jargon: some jargon is fine and even useful, but only when it’s well known in your field and useful to your audience. Avoid generic business jargon like “forward planning” (whoever does backward planning?), proactive (what’s the difference between proactive and active?), “leverage off” (if you mean take advantage of, build on or make use of, then use one of those phrases instead).
3. Headings: consider changing generic section headings in your document to express content. Instead of “Key Points”, call the section “Five reasons for switching providers.” This has two advantages: it is much easier to find this section in the document or in the table of contents; and it makes the document more interesting to read. Every document a manager has ever read has a section called “Key Points”. Yours will seem more interesting because it has “Five reasons…” instead.

Those are still all relatively simple fixes. It is in the body of your content where most of the extraneous matter will lie. To get at this, you need to step back from your writing and start by asking yourself two questions:

Why am I writing this? The answer needs to come in the form of a infinitive phrase i.e. to persuade managers to sign off a project business case; to inform the senior managers about progress on this project; to alert management to the risk of not acting.

Write down your answer on an index card or sticky note and refer to it as you read each paragraph in your document. Ask yourself, does this paragraph persuade, inform, alert or whatever your verb was; and does it bear directly on the object of your phrase i.e. the need in the business case, the project, the risk.

If not, remove the paragraph. You may decide later you do need it to explain or provide essential context, but take it out for now.

Next ask yourself,

If my reader only remembers three things from this report, what do I want them to remember? There is a good chance your reader will remember three things, but only if you choose them and make sure you get the message across. By choosing three things, you will at least determine which three things they remember.

Write the three on an index card and again, review each paragraph to determine which of the three that paragraph is contributing to. If the paragraph isn’t contributing to any, it may be diluting your message, and either needs to be removed, or reshaped so it supports one of your Three Things.

You may even decide at this point that you need to reorganise your document so that the tripartite emphasis is reflected in the presentation of your material.

The third question you need to ask is,

What is this paragraph about? Back in school you probably had something about “topic sentences” drummed into you. They’re still a useful idea. Look at each paragraph and ask, what is this paragraph about? Find the one sentence in the paragraph that makes your point, then delete half of the remaining sentences. If you have two sentences with two points to make (or worse, one sentence with two points), split them up. They are two different ideas, and need to be in separate paragraphs.

Sometimes this approach may get too bare-bones, and you may need to leave in some sentences to create a transition between paragraphs.

If you have done these things, your document may well be reduced by a third or even half. The final step is to polish your remaining sentences. Err on the side of short sentences. Why? Because your reader is busy, and short sentences are easier to read, they increase the likelihood that your point will come across clearly, and they significantly reduce the risk of ambiguity in your message.

The final polish is to clean up your terminology. Make sure you have been consistent in your use of specialist terms, and that you haven’t called the same idea, object or process different names in different parts of your document. See this earlier post on why consistent terminology matters.

Ridding your documents of clutter can be time-consuming, but it’s ultimately more time-consuming to write documents that no-one reads or wants to read.