If It Needs Instructions, It Doesn't Work

---

What’s even worse is when a product needs extensive directions but does not come with them.

But are you making a case for usability or a case against complexity? Things that do a lot may require more instruction than things that don’t do that much.

For web sites, the simpler the better. But when I’m flying from New York to L.A. I think I’d feel better about the well-trained pilots having as much control over the aircraft as possible – and the necessary level of control required to fly a jet of that size is inherently complex, even with the most modern of technology.

I don’t think the iPod needs much in the way of instruction (well, except for how to get music on the device in Windows, which isn’t as intuitive as “drag song files onto the new drive in Explorer”) but it still comes with detailed instructions that tell you how to press Play to have the music play, and so on. Jough Dempsey

---

In my experience, people don’t like to read the instructions anyway. As for me, only after some frustration in not figuring out whatever I’m working on, will I turn to the manual or help files. In that case, you’re on the money because there is “something wrong.” Joe

---

Amen. Usability in software has the ability to be self-understood. Instructions should (almost) never be used. The only place I can think of where instructions are useful is in Basecamp:

“Are you sure you want to delete this client? There IS NO UNDO.”....

And of course there is the old proverb: “Make it hard to screw up” JohnO

---

Even changing a light bulb (an extremely usable interface) can be made extraordinarily lengthy and complex when written out in step-by-step form. That doesn’t diminish the actual usability of the interface – it only acts as an additional line of support when the user gets confused, is unaware or unfamiliar with conventional wisdom and standards for similar products, etc.

Ideally, your interface or product should be highly usable with minimal or no consultation of the instructional text – for an “average” to “good” user. “Average” and “good” meaning different things for a light bulb vs. an e-commerce site vs. a passenger jet.

Unfortunately, not all users are “good” or even “average” – many users will be downright “poor”. Instructional text is written for the “poor” user test case.

There are many people that may be new to mp3 players in general, that need much more detailed instruction than others – education, if you will – that is necessary for them to even comprehend your product properly or begin to interace with your wonderfully usable interface.

There are people (and they do exist) who need step-by-step instructions for even the most usable and simple things. These are the people that highly detailed instructions are written for, as a last line of support for them to (hopefully) consult before hitting up a product support team.

Instructions, in many cases, are not a direct reflection on the usability of an interface, and are more a reflection on a company’s product support philosophy and practices. James

---

Another amen: you’ve summed up my most recurring problem in designing small sites for clients. We’re always searching for that balance between features, functionality and intuitive simplicity.

With web applications in particular, if a feature is getting too complicated it can often be slightly adjusted and simplified and still be as effective (accomplish the same task) for the end user. If the directions to use a feature are getting out of hand, but it’s the only way to have that feature, it’s time to see if that feature was really worth it to begin with. I’d say more often than not, it isn’t. Jeff Werner

---

To summarize my thoughts, I’m suggesting that the length of instructional text is inversely proportionate to the knowledge, experience, skill and intelligence of the absolute least common denominator that you are trying to service and support – not necessarily the usability of the product for the majority of its users. James

---

Garrett – I definitely agree that a solution should not be built to the lowest common denominator.

I also definitely agree that instructional text should not be used as a crutch to support poor usability for as many users as possible. That’s a valid concern and a symptom of certain cultures of design and development that needs to be addressed.

I only suggest that the instructional text should be written for the lowest common denominator that you actively define and choose to service and support.

Responsible identification of that lowest common denominator (and all of your users, for that matter), combined with professional instructional writing (and design and usability of the instructional materials themselves), and applied dedication to developing the most usable product/interface possible, for the widest gamut of users possible, will net you the right balance of usability and instruction. James

---

I guess it depends on the audience, the task you’re trying to complete and the frequency of said task.

Example 1: Searching for information on the web.

Searching the web should be easy because it’s a simple, repetitive task that need to be accomplished by a very wide range of people.

Example 2: Running a Nuclear Reactor.

The complexity of the task requires a complicated UI and a well trained operator.

While it’s true that the UI should be made as easy to use as possible, sometimes the complexity of the system means it cannot be used by everybody. Andy Budd

---

The examples of the airplane and reactor are interesting. While those are seemingly complex actions, shouldn’t they be NOT so complex? The process involved in both is not difficult. The operator, however, should be skilled. Flying an airplane is very basic. And the principles apply whether it’s a child’s toy or a new Airbus. But the operators require skills that make the outcome good, not bad. So in theory, shouldn’t even those actions be able to have at least some sort of basic instructions? I mean, on day one of pilot school, it can’t be that complicated…

Or maybe I’m way off base here. But I totally agree with what everyone has said here. I think Apple’s software falls on both sides of the fence here. I use my Mac all day, and 99% of the time I know how to get something done. But sometimes, the most intuitive task may require a trip to the help files. Those are the sorts of things that should have been caught before the product was released.

Thanks for bringing this up Garrett! It’s something that should be painted over the door of every design school… Bob Newill

---

If the fight is explicitly against inline instructional text (click here to learn how to use this form) that a majority of users are forced to rely on in order to accomplish something with a given tool, then I don’t think you’ll find many (if any) people in disagreement.

Your tool should stand alone, and function intuitively to every extent possible.

I think the debate may have come into the picture due to the generalization posed in the title of the article.

Beyond the fight against reliance on inline instructional text as a crutch to support a tool with poor usability, the issues of training, education, pre-existing user expectations, instructional design, complexity of task, etc., make the general issue a little more messy and difficult to sum up in such a definitive x=y formula.

Sparked a good conversation, though, so bravo. James

---

I find the definition of usability interesting:

The extent to which the device acts as the user perceived it should act. JohnO

---

I agree for the most part but there are some things where you fail to see the light. Some things should require instructions or else I don’t want people operating it.

Is taking drivers education not a set of instructions? I do not want people going out on a whim and assuming they know how to drive driving on the same road as I. Hell, I know people who have taken drivers ed and still can’t drive for crap (after several years, learn to not irritate me when driving).

And the number one thing and I know I am going to seem like a boarish jerk; true I am making an attempt at humor but I am being truthful as she sits right here. I wish my girlfriend came with a manual of some sort. I would have been familiar with the part where I do not say that as she looks somewhat irritated.

But as far as the overall gist of what you are saying; yeah I see it. Half the “toys” I buy weigh less than a pound and come with a manual that weighs more than the product. Ryan Latham

---

I like where you’re headed with this. As others have pointed out, however, some things do require extensive documentation. The question is whether excessive documentation is needed to explain the broken interface.

I’ve extended your idea a little further in my own post, with the Dimon-Messinger Usability Ratio. Adam Messinger