data collection for personas for a website for me, a website developer/designer

[johnyboy]

hello

i really like the idea of personas. the aspect i like about them the most: the help i think they'll provide in making early pivotal major decisions about a website, such as what content, what should the website do etc. -- real fundamental questions -- ones which are often made badly i think, almost skipped over and not thought about nearly enough sometimes.

i'm finally getting round to making my own site to get some website development/design work -- starting up my own freelance company. once i've got a website i can advertise (can't really advertise, as a web designer, without having a website to point to from the ads i don't think). this is a good opportunity to use personas. (a) to practice and learn about them, (b) i want to make a good site for myself and personas should help

i'm really stuck on the data collection part for the personas. i've got two good books on personas and they do talk about data collection, but i'm still stuck on it.

i can imagine, based heavily on real people i know or just know of who live locally to me, types of personas who are appropriate.

at the moment i'm thinking of segmenting on the amount of money they're prepared to spend, therefore the extent/fullness of the website they want/they'll get, and on the numbers of people the site will represent and the kind of organisation/venture

so at the moment i've got this segmentation (this is just an early assumption/sketch and i'm sure will need, at least, whittling down -- may even be entirely changed):

                       their budget for their site - extensiveness
                           | £x - £y (basic) | over £y (extensive)
---------------------------+-----------------+---------------------
a person with an idea      |                 |
(not primarily a profit    |                 |
making/driven idea)        |                 |
---------------------------+-----------------+---------------------
an artist.                 |                 |
their work                 |                 |
---------------------------+-----------------+---------------------
an entrepreneur. start up  |                 |
business venture/          |                 |
product/service            |                 |
---------------------------+-----------------+---------------------
one person business.       |                 |
company/product/service    |                 |
---------------------------+-----------------+---------------------
manager/owner of           |                 |
small (2-9 people)         |                 |
business (e.g. butchers)   |                 |
---------------------------+-----------------+---------------------
manager/owner of           |                 |
small (2-9 people)         |                 |
professional business      |                 |
(e.g. architects)          |                 |
---------------------------+-----------------+---------------------
a member of a small        |                 |
(2-9 people) non profit    |                 |
organisation               |                 |
(e.g. parish council)      |                 |
---------------------------+-----------------+---------------------
worker/middle-manager in   |                 |
medium sized (10+ people)  |                 |
company (same co. as below)|                 |
---------------------------+-----------------+---------------------
manager of                 |                 |
medium sized (10+ people)  |                 |
company (same co. as above)|                 |
---------------------------+-----------------+---------------------

some of those are based on the small number of people i've already done websites for, others who i know, others who i have imagined. (just for the record: me living very locally to the people who i've done sites for already has probably been *the* reason they chose me i think).

my main questions are, what kind of data should i be getting (further than the usual age, location, occupation etc. type of demographic info), and how can i get it? data appropriate for a web dev website? (*any* suggestions for any ways of tackling this would be appreciated)

if i were to interview some people (which i'm finding really hard imagining me doing, but anyway...) who fitted into some of the above, down the left hand side, categories what questions would i ask?! "if you were to want a website, what would be your main concerns?" ! the whole thing seems odd/off. i just can't imagine it happening. if someone hasn't considered getting a website then any quesions like "if you were to want a website..." are going to be answered flimsily at best. i can't even really imagine me interviewing people, apart from maybe in a non-official kind of happen to meet them, in conversation kind of way.

one book i've got talks about assumption personas, using assumptions to generate personas. this i can very much see me being able to do, but i'd really like to get and base them on some data.

i'm stuck. data collection for personas for a website for a website developer/designer? what? how? any ideas?

thanks.

[DCrux]

This is interesting, because it can be a content writing type of exercise. Most people try to "know who you're writing for." But they simply don't do it.

Personas are just as important for the copywriting of a web site as it is for the interface design of the site.

What's missing so far is motivations and objectives.

Where personas go wrong is their personalities don't fit their data profile. People try to force fit them into, say, being a willing customer for a web design company.

That assumption violates the basic rules of copywriting and marketing, not to mention persona design. The test most personas fails is they never make you uncomfortable. ....Sorry but that's not a persona -- that's an imaginary friend.

Your first step is to work under the assumption they don't want a website ...they want some objective a web site will help them accomplish.

A website just to have a website isn't an objective, it's an excuse.

Every persona has a name, picture, objective. The picture serves no other purpose than a last ditch attempt to get you to take seriously the first unhappy realization of interaction design: You Are Not The User.

It they don't want a website, what's the persona trying to accomplish? That's something you can find out through your interview process. A user who spends $X with you might spend $3X with someone else. Your assumption is that can't happen ...but that may be a faulty assumption.

The power of the persona is to question assumptions, not confirm them. Too many web designers use the persona process the way a drunk uses a lamp post -- support rather than illumination.

You want to illuminate, not work on your ventriloquism act.

Is Café Testing Right for You? Goes into what others don't: developing a test and writing a script.

At first you want to get into the habit of talking to people with an aim to get actionable information. Sorry asking people why they did business with you is almost useless: People Tell You What They Think You Want To hear. People who say they'll call you don't. People who say they're ready to buy don't. Live with it and learn.

You're trying to get good at listening to what users mean, not what they say. "It's not in the budget" may mean a lot of things, but it rarely means they might not spend five times the figure you're talking about with someone else. That someone else may deliver a website, but the actual client talk is about business building -- where the web site is a means to that end.

And it starts with the web design company site. Looked at from the user's perspective, they don't realize you spent two weeks on just the right stock photograph. Put it in a stack of ten or twenty other, identical, web design companies -- it looks the same.

Take local. Local is a piece of data, but doesn't explain anything. Why is local important for the persona? You have to get into the habit of asking why being local is important. Other companies, I assure you, have clients from all over the globe. Why? You had better plan on figuring out what is motivating the person who hires a non-local designer.

For a CSS website that validates, essentially a bucket for the content provided by the client, the budget is going to be X. The unpleasant realization is the persona went to TemplateMonster, and did a calculation where the result is X. That may be unpleasant. It may be "wrong" from your point of view -- that's the persona doing its job. Just count on every single existing client denying it. Nonetheless, when you're selling a CSS layout, the budget is going to be X. And the client is not going to budge.

For improving a profitability of a site where the last designer saw their business as providing a bucket for client supplied content, the budget may be 3X to 5X. You're doing the heavy lifting of developing a web business -- not a layout. You design a website explaining why pretty websites can make sales go down and why you fired all your graphic designers for visual merchandisers. Same thing for layout coding without the slightest idea of what the user interaction will be.

Second unhappy realization: You don't design user interfaces without users. The budget for UI designed without users is X.

And the result is a persona acting with a scenario where the client conversation needs to be entirely different to sell at higher prices -- and it's not just client demographics -- it's how and what you pitch. That's what makes this process fit within the content writing forum.

Don't be surprised if your data collection doesn't turn up a magic demographic who pays five times what you charge for doing exactly what you do now. If your persona is doing its job, that might be an uncomfortable realization. Congratulations, you've left the world of data collection and had your first fleeting brush with genuine information.

[johnyboy]

there's a lot in what you say for sure. i agree with much of it, and was already thinking along a number of the same lines, in particular:

> Your first step is to work under the assumption they don't want a website ...they want some objective a web site will help them accomplish.

i know that. (good book which talks about that: http://www.amazon.com/Think-Like-You.../dp/0071441883 )

[edit: although, by the time they're on a website which sells the service of making websites that's not such a correct thing. but the general point people don't want websites, they want whatever result a website might bring them is absolutely spot on. the author of the above book i linked to, his second book is titled 'selling results!' which, just the title, sums that up. in the book linked to above repeatedly there's a diagram: A -> B -> C. A is the current situation. C is where the client wants to be. and B is what you sell but of course you shouldn't sell B, you should sell C (and i suppose how B'll get them there)]

anyway, all this doesn't really help or address the original question. i'm not for a second saying what you've said is no good: i really do agree with nearly all of it, and as i say, already bearing it in mind. i want to not rely on assumptions and use data. that's what i'm stuck on. i'm really lacking in imagination on getting data/doing research to base personas on -- i don't care how they're segmented at the moment really. fine, segment them on motivation of getting website (primary one, which happens to be exactly the same as my own motiviation: get more/new work)

i really feel i'm asonishingly thick and lacking on this (this as in data and research). i'll probably just use assumption personas. i've got a real mental block on imagining how to get data for these personas. i do not have a mental block on imagining motivations and kinds of people at all though. clearly i'm destined to use assumption based personas whose only data driven aspect will be general, demographic kind of stuff rather than specific, related to the specific issue.

part of the problem, possibly, is what i want personas for isn't something people do/use regularly. if i was trying to get personas for a butcher's website for example, places to get data would be a bit more within my grasp i think. yes i think that's part of the problem possibly. there's certainly a problem somewhere that's for sure because i'm nearly completely blank on where to get (even what kind of) data from for what i want. so maybe a website for a web developer isn't an appropraite place to use data driven personas?

> Don't be surprised if your data collection doesn't turn up a magic demographic....

i haven't got any data collection! i'm a data collection invalid

thanks.

[DCrux]

Building a Data-Backed Persona do keep in mind this is the content writing section.

Usually it's not the data -- it's the collection. People welcome the data, they don't want to write a script and conduct a cafe test to go get the data. They'd rather chew off an appendage than leave the keyboard and face a real human.

Assumption personas can prompt data collection Assumption personas can be the eye-opening catalyst that gets your team interested in some real user research
-- The Persona Lifecycle

Come up with an assumption, devise a test, collect data which either proves or disproves the assumption. Maybe thinking about your assumption will spark a craftily constructed question you can ask during your next few client meetings.

Or you can call up clients who have paid $X thousands more than you charge and up for a web site and simply ask what they got for that kind of money. Then test what they say by putting it up and seeing who votes with their wallet.

One guy wanted to sell shaving mugs. That's manual shaving cream you whip up by hand. Nobody wants to buy that at retail (where he was selling it and going broke.)

What I had to do was some inventive data collection -- I went to a barber shop. They sell shaves. But they don't use "push button" shaving cream ...they too whip up their own foam by hand.

So I got into a chair and asked why. I didn't say I was doing research -- just asked. And the barber told me "When you're selling shaves customers want professional tools for shaving."

Insight received, I totally repositioned, repackaged, and reconfigured the product for barber shops and not retail.

I could have gathered web analytics about retail customers of shaving mugs for a decade and not turned up the obvious, but inconvenient, facts of the situation. As far as the client "knowing their business," sorry but this guy didn't. That's what I was there for, to find a business that kept his product out of a land fill.

What I'm talking about is extremely distasteful to a lot of people: customer and user research. People think that sort of thing drops out of a tree. The block comes in in thinking you're not going to have to go out, devise a test, and gather the data.

Character Development - Writers have to figure out characters a dozen times more complex than the average persona for development purposes. Maybe that's why the written content on the vast majority of web sites suffers ...developers are looking for the no-effort persona.

[johnyboy]

Assumption personas can prompt data collection Assumption personas can be the eye-opening catalyst that gets your team interested in some real user research
-- The Persona Lifecycle

Come up with an assumption, devise a test, collect data which either proves or disproves the assumption. Maybe thinking about your assumption will spark a craftily constructed question you can ask during your next few client meetings.

yup that sounds excellent advice. in your first response you were a bit negative towards assumption personas (or at least assumptions) possibly, and the Persona Lifecycle book is one of the the two persona books i've got and that book is, in one section in particular, very positive about assumption personas - places more importance and takes them more seriously than i would have thought. e.g. a case study of a guy using entirely quick assumption personas very successfully and various things dotted round the family planning chapter saying assumption personas, even though probably not as good as data driven ones, are much better than no personas - you still get a lot of persona benefits from assumption personas. then in addition of course there's what you point out: they can inspire and help with data collection for data driven personas (which is exactly what i need). yup, so that's great advice, thanks: press on with assumption personas seriously, and see where they lead, possibly giving direction on what data is missing, where to get it, what needs checking, how to check it etc. yup, that's what i'll do.

there's an excellent looking list of data sources in the persona lifecycle book (pg 124 - 126). in fact data collection for personas was the main reason why i got that book. i already had another persona book but thought the lifecycle one would cover data collection much better. it hasn't entirely satisfied me but has helped. that list (pg 124 - 126) is very geared towards already existing reasonable sized companies though. there's a number of books cited from the data collection section of persona lifecycle – for both qualitative and quantitative research. maybe some of those might be worth looking at.

What I'm talking about is extremely distasteful to a lot of people: customer and user research.

hmm, not to me really.

excellent, thanks very much for the info. very helpful.

off topic: do you like david ogilvy? i bet you do. he was a research fanatic. a little bit too much for my liking. if i was working for him i think i'd have ended up screeming at him "just trust your intuition for god's sake! you do have one don't you?". in particular that thought came from reading something written by him which went along the lines of: "i was very glad when humour in advertising was proven by statistics to work because up to then i had rejected many fantastically funny ad ideas". maybe it's only with the advantage of hind sight, but it seems obvious to me that humour in advertising has much benefit for making it work - to sell stuff. anyway. there was a great documentary on him on bbc tv just recently. very interesting. http://library.digiguide.com/lib/uk-...l+Mad+Man-3127 you could probably find it to download somewhere.

[Shyflower]

Building a Data-Backed Persona do keep in mind this is the content writing section.

Usually it's not the data -- it's the collection. People welcome the data, they don't want to write a script and conduct a cafe test to go get the data. They'd rather chew off an appendage than leave the keyboard and face a real human.



Come up with an assumption, devise a test, collect data which either proves or disproves the assumption. Maybe thinking about your assumption will spark a craftily constructed question you can ask during your next few client meetings.

Or you can call up clients who have paid $X thousands more than you charge and up for a web site and simply ask what they got for that kind of money. Then test what they say by putting it up and seeing who votes with their wallet.

One guy wanted to sell shaving mugs. That's manual shaving cream you whip up by hand. Nobody wants to buy that at retail (where he was selling it and going broke.)

What I had to do was some inventive data collection -- I went to a barber shop. They sell shaves. But they don't use "push button" shaving cream ...they too whip up their own foam by hand.

So I got into a chair and asked why. I didn't say I was doing research -- just asked. And the barber told me "When you're selling shaves customers want professional tools for shaving."

Insight received, I totally repositioned, repackaged, and reconfigured the product for barber shops and not retail.

I could have gathered web analytics about retail customers of shaving mugs for a decade and not turned up the obvious, but inconvenient, facts of the situation. As far as the client "knowing their business," sorry but this guy didn't. That's what I was there for, to find a business that kept his product out of a land fill.

What I'm talking about is extremely distasteful to a lot of people: customer and user research. People think that sort of thing drops out of a tree. The block comes in in thinking you're not going to have to go out, devise a test, and gather the data.

Character Development - Writers have to figure out characters a dozen times more complex than the average persona for development purposes. Maybe that's why the written content on the vast majority of web sites suffers ...developers are looking for the no-effort persona.

As I read the first part of this thread, I saw that something was missing, but I couldn't put my finger on it until I read DCrux's post above.

Notice, that DCrux didn't look for the persona of his client. He looked for the persona of his client's customer... the end user of his client's product or service who in this case turned out to be, not the person who got the shave, but the person who gave the shave.

Looking at your table, although the information in your first column is useful, it doesn't define a persona. It defines a business. The objectives of a lawyer, hanging out his shingle for the first time, differ exponentially from those of a new boutique owner or an inventor who wants to market a new type of shaving cream. Yet all three run small one-person businesses.

On your website, your job is to show your visitors how your web development company can meet their objectives. Before you can do that, you have to put a person into your persona, not a statistic.

[johnyboy]

Notice, that DCrux didn't look for the persona of his client. He looked for the persona of his client's customer... the end user of his client's product or service who in this case turned out to be, not the person who got the shave, but the person who gave the shave.

hmm, i'm not convinced. i know what you mean.

making personas for your client's client makes sense once you've got work from them, to help do work for them.

but then i know a client is going to be more receptive to someone who appears to have their client's concerns at the forefront of their minds.

i'm not convinced. a website advertising the services of a website designer is aimed at prospective clients (who in turn have clients who their website will be for). my prospective clients are the ones who will be using my website... ah, i don't know. most confusing...

as i say i'm not convinced. you have a point though.

dcrux was describing doing a job/work for a client, not describing doing work to get clients (promotional stuff). he was describing work that occured after the promotional work (if that's how he got the job) had done its work.

> Looking at your table, although the information in your first column is useful, it doesn't define a persona. It defines a business.

they are people not businesses. i know it doesn't in anyway go into the people but they are very much based on particular people not organisations. certainly in my mind they're people. i know that list isn't very descriptive. there's more in my mind for (most of) them -- not all i'll admit. mainly based on actual people who i know or at leas know of near where i live.


edit/addition: the people who will be using my site, who it's aimed at, are people who might want a site doing for them. my site should cater for people who might want a website. for example something everyone who's interested in a website want to know is 'how much will it cost me?'. that's a question any persona i come up for my site will likely have. if i was doing a persona for my prospective client's client, the 'how much will a webisite cost?' question is irrelevant and inappropriate. i conclude that the personas i make should definitely not be for my prospective client's clients but for my prospective clients. but those personas, all/any of them, a characteristic they'll have is a desire and keeness to help their clients, and if a website can do that for them, great, they'll like that.

[DCrux]

my prospective clients are the ones who will be using my website... ah, i don't know. most confusing...

This goes back to the beginning of the thread. Your task is the client's website. Your client's objective, in general, has something or another to do with their customers.

You client will be using your website. But the way they will be using it will depend on their view of web design.

In website design (your market, what you're doing the persona exercise for) has developers who are willing to give the client anything they ask for -- good and hard. One flash developer comment I read was their clients were idiots, but that if they were going to pay for stupid flash animation that lost them business, that designer didn't consider it their job to stop them from harming themselves.

Okay, that is one point of view amongst your competition. However, architects and interior designers (many, perhaps most) consider their job steering clients away from doing things they will regret later. The theory being someone who finds out after the fact the professional didn't exercise their judgment because they were too concerned about getting paid isn't worth doing business with a second time. Once burned, twice shy.

I believe some clients want code monkeys. They don't want a single shred of judgment or informed opinion, they want simple robotic execution. And other clients want the benefit of your good judgment and expertise about what the client's customers will respond to.

One persona is going to your site to see how robotic and compliant you are. They likely want to see a low price, and they want you to keep your design philosophy to yourself. They aren't paying you to think, certainly not about their customers. The attitude might be they know their business -- you don't.

The other persona is looking to see if you're the type that doesn't see your job as limited at building the container for their content, but to enhance their business. They want to see something beyond code or attractive layouts. They may see you as having worked on a competitor site. Or know something which could be used on their site from a similar industry that could give them a competitive advantage against competition.

Which do you imagine is going to pay rock bottom price. And who, do you suppose, is going to pay a premium price? If you can find some tool to step outside your business, you may find the answer.

Typically, developers don't like to segment their market. They're looking for the lowest common denominator technology which is everything to anybody, on the assumption they'll get all the business. Inventors do this. Technologists do this. And it rarely works for any kind of value added business. If you're not doing the stuff "they pay the big bucks for" you don't get the big bucks.

My opinion on assumption personas: They get better the more persona designs you do. The more personas you have done, the better your judgment is going to be. You're not relying on a specific data set, but your experience from a whole lot of varied data collection and experience with human nature. They aren't assumptions, really, but based on qualitative data rather than quanta.

So, if your background has something to do with human observation, like an anthropologist, then do assumption personas. If your time is spent avoiding human nature as chaotic and frustrating compared to code, then you shouldn't be making assumptions about human nature -- not without quite a few data backed personas under your belt.

I think assumption personas are good for people who have developed the discipline of regular persona creation. If you are writing a story and have developed a lot of well researched believable characters, you can "wing it." If not, then not. If you have a writing background of doing this, then your assumption personas are probably going to be pretty good. If you don't have a background which would lend itself to the "you are not the user" discipline required, not to forget storytelling, I can see disaster ahead.

Related:

The legend of how J.K. Rowling wrote Harry Potter on scraps of paper reminds me of the agile development technique called the Wall of Wonder. The basic objective is not to let isolated scraps of data stay as unconnected. Personas can be seen as a big picture framework for processing the data scraps into information.

A lot of developers say their users only used 10% of the app, but each uses a different ten percent. Really, they may be looking at five or ten different personas using lowest common denominator technology to accomplish widely differing goals. a Wall of Wonder can help you see where you're lumping different customer segments. And it also helps when you're writing something on the order of Lord of the Rings or Harry Potter, with a lot of plot strings to keep track of.

[johnyboy]

my prospective clients are the ones who will be using my website... ah, i don't know. most confusing...

This goes back to the beginning of the thread. Your task is the client's website. Your client's objective, in general, has something or another to do with their customers.

i'm not actually that confused on that. i'm sure the personas should be for the people who'll use my site, and personas representing those people have, as you say, an objective to please/cater-for/help their customers -- that's an attribute (a very important one) of the personas for my site, but not to make into personas in their own right. i definitely shouldn't be making personas for people-who-might-want-a-site's customers. i definitely should be thinking about appealing to personas representing people who'll use my site, and one way to do that is via the promise/demonstration/whatever of pleasing their customers via their yet to be site in some shape or form.

all this 'customer of web designer wants/needs to please their customers' aspect is definitely an important thing. i was already very much thinking along those lines before writing this question. honest. i really was. some of the preliminary pages i've sketched, just as thoughts, are all based on things which are good for my customer's customers. e.g. 'useful'. that's one theme of a page. useful websites to my customer's customers that is. rather than a page based on say 'php programming' which is a feature about a website. thing is i was already thinking along those lines before thinking of using personas. it's not personas at all that's helped me think of making a webpage based on 'useful' rather than 'php programming'. i hope personas'll help extend/improve that kind of thinking though possibly.

My opinion on assumption personas: They get better the more persona designs you do. The more personas you have done, the better your judgment is going to be. You're not relying on a specific data set, but your experience from a whole lot of varied data collection and experience with human nature. They aren't assumptions, really, but based on qualitative data rather than quanta.

i'm sure you're right but i'm looking to use assumption personas as a placeholder, a way to start, a sketch, an in road to doing data driven personas. i think they're going to help me do data driven ones. i'm seeing assumption personas as an aid to data driven ones, not as personas to use as are. that seems reasonable to me. as a way to help get research ideas -- which is what i am (or was if assumption personas do what i hope they will) stuck on.

a Wall of Wonder can help you see where you're lumping different customer segments. And it also helps when you're writing something on the order of Lord of the Rings or Harry Potter, with a lot of plot strings to keep track of.

hmm, yup i like that. interesting.

some excellent info in general in that reply dcrux, thanks very much.

[Shyflower]

My point is before you start building personas, choose a niche market. What is the focus of your business?

Again DCrux makes excellent points. You can't be everything to everyone, but on the Internet everyone has a chance to see your site.

I think everyone knows that some businesses have 1 to 10 employees, some businesses have hundreds, even thousands of employees, and other businesses fall somewhere in between. What group can best use your services?

Choose one and then further define your market. If it's the micro-business group (1 to 10), who are your clients. Are they lawyers, plumbers, or inventors?

Choose, for instance lawyers: Are they corporate, civil, criminal or general practitioners. Will you target established firms or startups. IMO, it's at this point you start researching and building personas.

Or perhaps, you choose a larger group, such as all small retail businesses in Podunk USA.

Define your market first so that you can develop personas within your market.

[johnyboy]

> My point is before you start building personas, choose a niche market.

i see. i thought you were saying make personas for my prospective client's clients.

> Again DCrux makes excellent points

agreed.

> ...
> Define your market first so that you can develop personas within your market.

right i see. got you. yup, was heading that way. the empty boxes in the table in my first post are where the actual personas reside and i know it all needs whittling down in numerous way.

ok, thanks.