Thursday, September 29, 2011

Learning to read the world through other eyes

Here is something a friend sent me this morning that I have already scanned once and once is enough to make me want to share. It's about thinking tools; it's a little too academic; it includes text such as this:

As it is often difficult to notice or examine the hands that are writing us, TOE was designed so that learners can develop  tools to enable them  to notice and analyse these hands more easily

Please read and let me know what you think.


Thursday, September 15, 2011

EthOS check list

Here's the deal: you have just signed up and downloaded the app. And about to start your first ever EthOS ethnography project. And I stress ethnography project, which is what this check list is for:

Creating your project on EthOS
1) Examine your project objectives/deliverables carefully. Some will be observation based and some will be question based. Do not mix them up!

2) What do you absolutely need to capture (film & photos) to help you meet your observational objectives?

3) What do you need to ask/probe (text & audio) to meet your question based objectives?

4) When inviting respondents (we call them researchers) in 'manage project people', make sure you tell them in advance that they will be receiving emails with QR codes to scan and launch their project on the EthOS app. We receive many support emails from respondents asking us not to spam them!!! Because they have forgotten they are taking part in a project and they had no idea they were going to receive a QR code to scan.

5) Know your permission levels when you invite anyone to your project. Note that a project manager can invite other project managers without your permission.

Recruiting respondents (people)
1) No matter how carefully you need recruit, don't ignore sites like Craigslist to find respondents. We have used it successfully on a number of occasions as have a bunch of clients. The only issue is the large drop out rate after respondents have agreed to take part. Half may not show up! So be sure to double up your sample.

2) If using a recruitment shop - your drop out rate will be far less, but your costs, including incentive costs, will be much higher. However, it will save you a lot of valuable time and potential headaches with having to deal with recruitment yourself.

3) Once the project begins, expect 1/10 respondents to not contribute often or effectively enough. Careful recruitment will minimise this. BTW, this figure is 2/3 for discussion groups.

4) When recruiting, be clear that all household members must be happy with participating in your study. Even if they are not directly involved. It can be annoying knowing that family members (usually dads in my experience) are hiding just off camera.

5) If you are going to be using them for additional explorations - we sometimes do our own participant observation with people at the end of a journaling phase - make sure you warn them before the project begins. Nothing worse that a client asking to deep dive a particular respondent and finding they don't want to be paid a visit!

6) Decide how much content you need. Are you making a film for your client as an output? How many video entries a day would you need. We suggest, for a typical project (I know there is no such thing), 3-4 entries a day with 1-2 being video entries. Any more and people may start thinking too hard about what they can capture for you. Killing any semblance of naturalism.

Generating themes (tasks and questions)

1) If you don't want to bias respondents towards a particular, brand, service or activity you will need to disguise it. Disguise by stating additional brands/categories that you need them to capture for you.

2) Have a mix of continuous themes - across the duration of the study and one-off themes to keep respondents engaged. Note you can change themes easily in your project, 'manage project themes' and the changes will sync automatically on every respondent's device.

3) We recommend a maximum of 10 and ideally 6 themes at any one time. These can be tasks and questions. But keep them to one (seven word) sentence long where possible. You can also use a single word, e.g 'snacking'. (Note people will be reading your themes on their devices (against which to capture their entries) and the fewer lines to read the better.)

4) Themes are locked down - respondents cannot add to them. So we suggest always having a additional theme called 'other'. This way, if they want to send an entry which doesn't obviously match one of your themes, they can still send it to a theme bucket, albeit a general one.

5) Tags can be created by respondents and should be used to add detail/context to the themes they have selected.

6) With both themes and tags it is fine to select more than one from each.

Final check before launch
1) Be sure to allow yourself a day or two between being ready to start your project and actually starting your project. Use this time to ask each and every respondent to send you a few entries. You might even set a simple introductory task.

2) Sometimes people don't realise how weak their Wifi is until they try sending a video. It's worth identifying and troubleshooting these people before the project starts.

Post fieldwork

1) This is a great time for deep dives with the best of the rest. Which ever 'online researchers' you are going to visit, delete their emails in 'manage project people' and add them again as 'offline researchers'. This way you can use your device to capture entries of the respondents, as the respondents themselves. Confused? All this means is that you can still filter by respondent names and see your entries of them under their names.

Please feel free to call any of us if you would like help with any stage of your project. Our aim is your 100% success!


Tuesday, September 13, 2011

We have permission to reproduce

I am reproducing, below, an anonymised evaluation report. It was for a client who commissioned a study using our platform. The agency is one of the most globally respected. The client for whom the report was written is a household brand.

We have permission to reproduce this:

"Research on Research

This was the first project xxxxx have run using Ethos (apart from a small internal prototype). Overall we feel it was very successful and is able to create exciting new insights for brands. It is a unique methodology that covers both qualitative and ethnographic styles of research.
This methodology has a number of potential applications:
• Identify different beliefs and motivations towards a category
• Gain insight around routines and rituals
• Build hypothesis for segment creation
• Provide colour and context for existing segments
• Uncover role of family dynamics
• Understand difference between expressed belief and actual behaviour

What worked well

Unknown unknowns
Creating broad tasks for respondents leaves room for the unexpected to reveal itself. Traditionally research is confined by the expectations of the researcher, using grounded theory analysis lets us move beyond this.

We purposely did not read the client segmentation work before the analysis so were not led in our analysis. Despite this, we identified eight of the nine need states (although they were categorised slightly differently).

Observable routines and rituals
We can see not only how often people treat their dog but also what happens when they do. For example, whether they use treating as an opportunity to play, train or simply to derive pleasure from seeing a happy dog.

Shows up relationship dynamics
Most dogs exist in a family, this creates its own family dynamics. For example the role children play in treating – typically looking for the most excitable treat, asking the parent’s permission before treating. Partners can act as treating ‘handbrakes’ – dissuading the other person from excessive treating.

Human findings
As the data mostly consists of pictures and videos the story it tells is an extremely human one. This makes it a far more intuitive way of learning about people’s habits, beliefs and behaviours.

The fieldwork is ten days which provides a good snapshot of people’s lives including both typical and non-typical days. It also provides the opportunity to digest what they’ve said and probe further where necessary, this allows you to focus in on interesting aspects of their lives. There were 475 pieces of content uploaded, from 12 people over a 10 day period.

Unobtrusive technology
People feel very comfortable using their phones as a capture device and this allowed a real window into people’s homes. Although there were a few technical problems during the first task, these were resolved by the second. Some of the participants asked to be included in future studies.


Self Vs other ethnography
The nature of this project was very home-focussed which made it difficult to replicate traditional ethnography where they would actually go and observe other people in their homes. We did ask them to interview a friend about their dog, but this was more of a structured interview rather than an ethnographic situation. Some categories may be better suited to a methodology where the participants become the ethnographers, this could be achieved but would require more incentives and potentially some training.

From end of fieldwork to final report took around a month. This was partly due to getting to grips with a new approach but also the depth of the analysis, in particular creating the prezis (the initial word-based analysis took under a week). In future more could be done during the fieldwork to prepare the prezis and the final report which would speed this process up, developments to the ethos site site also will make this a smoother process for the researcher.

Some of the events felt staged for camera, this could be because participants were asked to upload six items per day. Even in these occasions there was still a lot for us to learn as the owner’s still maintained their standard treating ritual. To get more authentic occasions we would need to make sure that the tasks fit naturally into their lives (e.g. photograph every meal you eat).

Only 2 of the active 12 participants had children in the house, would have been better to get some more households with children. Out of the 25 people we invited to take part 10 were households with children, it could be that we were unlucky to get a low response rate from families with children, or that these families have less time to conduct this type of study. Next time if children in the house is an important consideration we should aim to over-recruit slightly more.

It is hard to record yourself, especially during a treating moment, holding a camera will limit your actions. It worked well when people got other family members to video them, or the participant was videoing someone else. It might be useful in future to recruit families to partake as a unit of analysis rather than an individual, although this may pose some technological problems (i.e. they all need iPhones or androids), we could get round this by giving them a family flip camera for duration of the fieldwork. It would be interesting to get multiple perspectives in this way."

I do have a few comments to add - especially about people setting up events and challenges recording ones self. The best way to avoid is to ask people not to! And not to set too many tasks. We recommend around 3 per day.

In India we used the platform for a study involving people filming their own routines. People managed with help from friends, family, props and even an old cassette case which worked beautifully! But perhaps we should send people those cheap 'n cheerful iPhone stands...

Would value feedback from readers of this post.

Wednesday, September 7, 2011

Talking shop

Running a mobile research platform isn't all that!

Go get a hot drink, kick back and find out why.


Monday, September 5, 2011

From hairdressers to persian carpet sellers

Ever heard about the hairdresser who became so successful he had no time to cut his own hair? He stopped be abe to see what he was doing!

One of the things (there are many and this is just one) I wish I could do is an ethnography of everyone using our app. We do ethnography's for others, but never for ourselves. I'd love to see how users design tasks and questions around their project objectives. I'd love to see how they use the features and workspaces in particular. Most interestingly, I'd love to understand how an organisation which has a long tradition of working in a particular way suddenly adapts to our new platform. What would happen if you imposed our app on an organisation?

Let me give you an example. Remember this article? Watch the short film before reading on... And note the app is a very old version.

See the guy in it? That's Bernabe. Arguably one of the world's most talented and experienced video ethnographers. He's my right hand man - no he is both hands. And he manages most of our projects from beginning to end. When time allows, he actually loves to film in-home himself. But typically, he is managing a bunch of other ethnographers in different markets.

Reason why I mention him is this: When I completed the prototype app, I called him and asked him to give it a whirl by downloading it to his iPhone. Note up to this point, Bernabe had little or no involvement in the  development of the app. A few days later he called me up for a chat:

"Remind me, why did you create this app?"
"Well, you can now come back from the field with all your films ready in your inbox" I replied patiently.
"But it really won't save us any time if that's what you're thinking."
"Bernabe, you spend days endcoding video and then clustering the themes together." I countered.
"Yes but most of our time is taken up making films - not just clustering and theming"

And so we continued arguing about just how useful an app would be.

Bernabe didn't use it. I couldn't force him to. And there was no way he was going to select our ethnographers on the basis of whether or not they owed iPhones.

Not long after Bernabe's unsuccessful pilot of the app I called him for another favour. To ask him of he could make a film of the App in use. Like a demo film to help people understand how I see the app being used. He had no objections. And we selected a tame household to make our film with.

Apart from calling me at the end of the filming day to say the app crashed about 200 times, he said nothing else against it. And this, I believe, was a turning point.

Before long, we were using our iPhones in the field along with our PD150's as back up video cameras. And not long after that, people were using Blackberry's and Android versions of our app too.

A few points to note:

Ethnographers were using the app to do participant observation before we even realised it might also be used for journaling. It was a genuine surprise to learn we might need to build a 'lite' app just for respondents (still building it).

There was no big launch of the app within our organisation. No one was forced to use it. And when they did, we would hear lots of comments and helpful suggestions from the users while often still in the field.

We made no promises with the app so as not to break any promises. It was very much a soft launch. Discover it for yourself we said (and still say). Your frst project is free. We did also persuade them to include us in their projects so we could trouble shoot and help out rapidly when needed. Soft launch or not, we want everyone to have a successful experience.

As we added new features (I'll blog on this soon), we stuck to the simple principle that we were building a tool for ourselves. Not anyone else. We weren't trying to make a fast buck by appealing to a large audience. This was our tool, designed to be used with our own methodology/process. If you like it, fab! If not, find another app or build your own. Seriously.

And when we realised that we did need to use the app among respondents who were journaling and keeping diaries, we started to think hard about a larger platform and not just an app. And I can now tell you that building this platform will never stop.

Going back to my hairdresser analogy, we are now beginning to cut our fringes back a little so we can see what's going on. We have a new commercial director - ex BrainJuicer - to help us with that. And what's going on is that we are selling people Persian carpets. Ever been sold a persian carpet by an Iranian carpet seller? They will suggest you take the one you like home and throw it on the floor to test for a few days or weeks. If you like it, pay for it. If you don't, bring it back in reasonable shape. After all, the more you walk on a Persian carpet, the better it looks.