Redesigning the Mind website

In 2013 we redesigned the Mind website, to make it easier for people to use.

Lots of people told us that the old website was hard to use, and sometimes overwhelming. We wanted to make our website easier to use, so that people can seek our help sooner, get behind our campaigns, and support us more generously.

We’ve tried to base our new site around people’s needs, so the design process was shaped by over 50 users at in-person workshops and user testing, and by over 100 people who participated online. We held workshops, surveys/tests, and user testing.


We held workshops so that we could understand what people need from the Mind website. We wanted to find out what their top tasks were, and asked people to devise a site structure and labelling system that made sense to them, using card sorting. The aim of this was to make a site structure that isn’t based on organisational language or the way that we structure our work or teams internally.

We also invited people to discuss their feelings on the Mind brand, imagery, and use of colour online, and their feelings on how Mind’s values apply to the website.

We held the workshops in Bristol, London, and online.

People’s top tasks were:

  1. 1) Finding information about mental health
  2. 2) Knowing that I’m not alone
  3. 3) Finding support/services near me
  4. 4) Getting involved (mostly with fundraising)

People told us that they want a greater focus on their most important tasks. They wanted less clutter, a calmer and more sparing visual design, with a greater focus on people. They really resonated with Mind’s values: real, personal, compassionate, courageous.

Notes from a discussion about Mind's brand values and their relevant to Mind's website

Using this information alongside other information about our web visitors – eg from Google Analytics – we were able to draw up a clear, shared understanding of what our project was trying to achieve, and what the most important tasks were. This was crucial, as design is about prioritisation, and that often means telling people that their area is less important. Being upfront about your top tasks – and getting these tasks from users – helps with these discussions.

The in-person workshops took about 4 hours; the online workshop was condensed into an hour.

Our understanding of top tasks also drew from our Analytics data, which tells us things like over 80% of our traffic is to our information pages.

Online Survey

At the workshops, participants devised a site structure that made sense for them. We wanted to test if it made sense to other people too, so we tested it.

We carried out an online reverse card sort exercise to do this, using a tool called Treejack.

There were some areas of disagreement between the workshops (eg “get involved” vs “support us”, and “info & support” vs “help & advice”) and this exercise helped us decide between the competing options.

The exercise took each user about 10 minutes.

User Testing

Before we started building the final website, we built a sort of draft version: a prototype. It was just made with HTML and CSS. We did 1:1 user testing and online user testing of this.

By asking people to carry out a number of top tasks, and seeing what they found easy or hard, we were able to improve the prototype before building the final product.

We did this as part of following an agile methodology.

We encouraged participants to think aloud as they went so that we could better understand their thought processes. We took notes and for one of the sessions participants agreed to have their screen and faces recorded, using a tool called Silverback.

user testing the mind prototype

Some key learning points

– I’d start the recruitment process earlier, and plan a whole journey of engagement, rather than just inviting individuals to one-off events.

– We had more people sign up than we expected or had capacity for. (This did help us select a more diverse group for our London workshops.) I’d have liked to plan even more opportunities for wider engagement.

How did all this user involvement help us and help the people who took part?

By better understanding our visitors and their needs, we built a better website for them. By testing as we went, we could make sure that our designs were on track.

Participants gave very positive feedback, saying that they felt listened to, valued, and that their feelings and ideas were taken seriously.

What support did you offer?

– We tried to make sure everyone felt comfortable at workshops. We made sure that we had refreshments, a quiet space for people to go if they wanted a break, and we also asked people to fill in a wellbeing plan so that we could know how best to help them if they were distressed. (Most people didn’t fill in a wellbeing plan, but some did and it was very helpful in these cases.)

– We used supportive language and set an expectation of non-judgmental listening and sharing. We made sure not to ask leading questions, so that our results were useful.

– We tried to be clear in advance about what people were signing up for, and we told people that they could withdraw from activities at any time.

Tips for other people doing user involvement activities

– Be clear in your aims. For us, the value was clear: good agile digital project work relies on user engagement, so doing lots of user engagement was always a priority for achieving a good result. Mind takes user involvement seriously, and this made working in an agile, user-centred way very easy to achieve.

– Budget the time and money to carry out engagement activities. If you want the benefits, you need to plan the investment.

– Negotiate administrative support early, to help with booking venues, paying people for participation, etc.

– I pre=scripted everything non-reactionary in the online chat. i.e. all the welcome messages and pre-planned questions were written in advance so I could copy-paste. That left me free to focus on responding to the contributions of participants.

Best Practices & Lessons Learned from 30,000 A/B and Multivariate Tests – summary of Optimizely’s video

Optimizely’s Best Practices & Lessons Learned from 30,000 A/B and Multivariate Tests video is an advert for A/B testing in general, and for Optimizely Optimizely as a tool for running and creating tests.

This video features some useful examples of multivariate tests done during the Obama campaign. Here’s my summary:

  1. Nothing is sacred. Question assumptions
  2. Explore before you refine
  3. Less is more
    Some content experiments: – One Kings Lane: Removing 2 fields led to an 8.4% increase in signups
    – Clinton Bush Haiti Fund: Removing title field and phone number field saw an 10.9% increase in the $value of each visitor’s donation
    – AVG Australia: Removing links from the checkout process increased the $ value of each visitor by 16.8%
  4. Words matter. Focus on your call to action.
    – Donate Now, Please Donate, Why Donate?, Donate and Get a Gift, Contribute: which is better? The status of the individual makes a difference to which of these options is most effective: new people respond best to “Donate and Get a Gift” (+15.2% $ per pageview compared to Donate Now); people who have signed up but never donated responded best to “Please Donate” (+27.8%), and people who have previously donated responded best to “Contribute” (+18.4%)
    – Clinton Bush Haiti Fund’s donate form – changing the confirmation button from “Submit” to “Support Haiti” increased $ per visitor by 15.75%.
  5. Fail Fast
    You can get statistically significant results very quickly.
  6. Start Today
  • Q: How many variations should you test at once?
    Tradeoff between breadth of test and speed of reaching statistical significance.
    A good starting point is to test with 4 variations, then take the top 2 and test them again.
  • Q: What percentage of website traffic should you allocate to a test?
    You can control this per experiment.
    They recommend allocating as much as you can, to reach statistical significance sooner.
  • Q: Can you target specific types of visitors?
    You can choose to target all, new, or returning visitors.
  • Q: How do you target paid traffic?
    Do this with the targeting interface: trigger by origin location, or match certain query parameters (eg utm_source)
  • Q: How do you test dynamic pages?
    Go to interactive mode to make the dynamic stuff happen.
    Tell us when you want to activate optimizely. By default it loads on page load; but you can activate it manually through the API.
  • Q: Does Optimizely work for pages behind logins or paywalls?
    Yes, as long as you’ve added the snippet to the page.
  • Q: How do you test big changes to a page, or test a new page against an old one?
    If you want to test a big change to a page, or test against an old page, you can make it so that visitors selected for a specific variation are redirected to another URL. (Make sure that the snippet is on both pages)
  • Q: How do you test server-side changes? i.e. where the variations are reflected on the backend and not just on the front end (eg graphics). Eg price variations.
    A: Optimizely wants to improve at this. Most people doing this currently set up two different pages, and set up a javascript gate on the front end to pass 50% of visitors to one page, the other 50% go to the other.
    See and
  • Q: How soon to pull the plug on a variation when it looks like it’s not performing well?
    A: Wait until you get statistical significance. The early days of at test may not be truly representative.

Twitter for good – 7 November 2014

On Friday 7 November, Twitter UK hosted ‘Twitter for Good’ at their London offices.

There was an initial presentation aimed at a broad group, presumably ranging from people who don’t use twitter through to charities making heavy use of the platform. This meant that there was a good range of material explaining and selling the platform as a whole. There were also some good statistics on the behaviour (and, implicitly, value) of twitter users to charities:

Propensity of twitter users to give

Taken together, these figures suggest that the interest in fundraising content is higher if the ask or content is right.

Importance of mobile

For anyone making the case for a responsive website, there was a great stat on mobile use of twitter:

Twitter cards

Twitter cards can be used for signups too:

We’re currently thinking about Data Protection, so I wonder how easy it would be to integrate twitter cards into our overall data capture setup.

This event signals the intent of twitter to deepen engagement with charities

Twitter’s interest is to deepen an extend use of its platform – and, implicitly, to monetise this use. To my surprise, adverts weren’t mentioned at all. I guess growing an engaged user base is the key challenge; advertising comes second.

I’m looking forward to what may happen next. Could twitter operate as a hub for charities sharing best practice, and for condensing and disseminating these practices back to the twitter-using charity community? There’s clear mutual benefit there. Hopefully we’ll see some developments in the new year.

Why I disagree with GamerGate

I oppose GamerGate, and I’d like to explain why. Please be aware that much of what follows is unpleasant and NSFW.

GamerGate began on unsound foundations

GamerGate started off with Eron Gjoni posting an angry rant about his ex-girlfriend Zoe Quinn. Zoe Quinn was accused of using a sexual relationship with journalist Nathan Grayson to obtain favourable coverage for her work.

This claim isn’t convincing: Grayson never wrote a review of Quinn’s work, and the article that he did write which mentioned her work was published before they were in a relationship.

But even if these claims were convincing, GamerGaters targeted the wrong person. Grayson was the journalist, yet Quinn was the target of vitriol. If GamerGate is a movement about ethics in games journalism, I would have expected it to focus on the journalist.

I don’t think that policing the sexual conduct of one female indie game developer is the best way to make the case for ethical improvements to games journalism. If you are concerned with unethical actions of games producers or publishers, I think it’s much more useful and courageous to focus not on minor indie developers, but on big AAA publishers. These are the companies who have the power to undermine ethical games journalism, and who actually have a track record in doing so.

There are real, known ethical issues in games journalism, but GamerGate doesn’t seem to be focusing on them

GamerGate did not arise in response to the following ethical breaches:

  • A retail chain owning a gaming publication. (The same company owns Game Informer magazine and Game Stop games stores.)
  • Journalists being fired for standing up to the demands of a publisher. (Jeff Gertsmann being sacked from GameSpot after posting a negative review of Kane & Lynch)
  • Product placement in Youtube reviews. (Geoff Keighley being interviewed by Pixel Perfect with product placements for Halo 4, Doritos and Mountain Dew)
    Geoff Keighley product placement
  • PR companies controlling the editorial output of reviews. (Plaid Social requiring Youtube reviewers of Middle Earth Shadows of Mordor to submit their reviews for approval before publication.)
  • Certain publications being granted exclusive early preview access, incentivizing them to produce positive copy. (Eg IGN First.)
  • (See also Leigh Alexander’s list of ethical concerns that have not led to furious activity.)

Out of concern for journalistic ethics, GamerGate has challenged journalists using Patreon, which has led to Kotaku banning its journalists from using Patreon, and to Polygon requiring staff to disclose Patreon use.

But this relatively small area of attack doesn’t address the above concerns. Nor does it identify or seek to challenge the key conflict of ethics at the heart of games journalism. As this article published on Deadspin puts it: “From the top down, publishers ranging from AAA behemoths like Electronic Arts to the IndieCade crowd do in fact enjoy symbiotic relationships with gaming media outlets, and if it came down to nothing more than sex and petty corruption, that would be nice, because the problem would certainly be a lot more easily solved.” In short, games companies have a vested interest in supporting journalists to promote their games. Changing Patreon policy in a few places doesn’t challenge this core problem.

So far I haven’t seen GamerGate seriously engage with ethical questions in games journalism. But I have seen GamerGate actively working to attack journalistic ethics.

GamerGate’s main campaigning effort directly opposes journalistic ethics

Gamergaters have attempted to punish sites that they disagree with, by acting to remove their advertising revenue. As a result of Operation Disrespectful Nod, Gamasutra lost advertising from Intel; Gawker lost advertising BMW, Mercedes (later reinstated, I believe) and Adobe.

If GamerGaters are unhappy that the press is not sufficiently independent and principled, attempting to influence publications by lobbying advertisers to withdraw their adverts seems like an odd move. The aim of Operation Disrespectful Nod appears to be to make the press less independent and to intimidate journalists who hold opinions that GamerGate supporters disagree with. In other words, this course of action suggests that GamerGate is opposed to ethics in games journalism.

I want to look at sexism now, as it’s a dominant theme of GamerGate.

GamerGate is an opportunity to attack women

GamerGate coincides with a wave of threats of violence against women. Zoe Quinn, Anita Sarkeesian and Brianna Wu have recently left their homes in response to threats of violence.

Some people have dismissed these threats as publicity stunts: e.g. “Let’s be honest. We’re all used to feeling a niggling suspicion that “death threats” sent to female agitators aren’t all they’re cracked up to be. And indeed there is no evidence that any violent threat against a prominent female figure in the media or technology industry has ever been credible.”

I strongly disagree with this. If someone is making something up they probably wouldn’t go to the FBI with evidence. And violence against women is a big problem, and not just something that people make up on the internet to get attention. In Britain, more than a million women experience domestic violence each year. The UK government reports that approximately 85,000 women are raped on average in England and Wales every year, that over 400,000 women are sexually assaulted each year, and that 1 in 5 women (aged 16 – 59) has experienced some form of sexual violence since the age of 16. We need to take threats of rape and violence seriously, not as the currency or cost of online discourse.

Some people have said that the women being attacked online have brought things upon themselves. e.g. “goading people into making unpleasant remarks and then using those statements to publicly beg for sympathy and cash,” or “They’ve become professional victims.” I think this is incorrect. I haven’t seen any deliberately inflammatory material put out by Sarkeesian or Quinn prior to their being attacked. Rather, I think that their main ‘crime’ has been speaking about and analysing games. Regardless, nothing justifies rape or death threats. To say that Quinn, Sarkeesian etc have brought things on themselves is to harness a common misdirection in our culture: blaming the victim of a crime rather than the perpetrator. In GamerGate as with society more generally, it’s wrong.

Anita Sarkeesian should have nothing to do with GamerGate. Her Kickstarter-funded project Tropes vs Women in Video Games is an example of open, clean funding for independent journalism. But Sarkeesian has been under attack since before GamerGate, and things have intensified:

Women engaging with GamerGate in a cautious, neutral fashion are being attacked. Felicia Day aired her concerns about being attacked if she shared her views on GamerGate: “I realized my silence on the issue was not motivated by some grand strategy, but out of fear that the issue has created about speaking out… I am terrified to be doxxed [having your public contact details / address shared] for even typing the words “Gamer Gate”. I have had stalkers and restraining orders issued in the past, I have had people show up on my doorstep when my personal information was HARD to get.” Day’s article included passages like “Games are beautiful, they are creative, they are worlds to immerse yourself in. They are art. And they are worth fighting for, even if the atmosphere is ugly right now.” She was doxxed shortly after publishing this post.

Jennifer Allaway was working on a study to understand “the importance of diversity in game content to game players, and whether or not the game industry is able to predict this desire.” Gamergaters discovered her survey and spammed it with loads of sexist entries. “if you’re even asking about equality or diversity in games, being shouted down in a traumatizing manner is now a mandatory step that you have to sit back and endure.” “Had #Gamergate participated in my survey honestly, as a researcher, I would gladly have taken their data… But instead, #Gamergate left me with hundreds of replies consisting of bald-faced mockery and threats.”

But men are safe to participate in online discussions about GamerGate, even if totally inflammatory.
Chris Kluwe, former American Football player, wrote an article on Why #Gamergaters Piss Me The F*** Off, in which he called GamerGaters “Basement-dwelling, cheetos-huffing, poopsock-sniffing douchepistols”, and a load of other things, and did not get doxxed.

I haven’t seen sexualised, violent abuse and doxxing levelled against JonTron and Adam Baldwin, prominent figures in the pro-GamerGate movement. This suggests that there is a problem with the treatment of women in GamerGate. As such, I was not convinced when Huffington Post Live asked two women who support GamerGate whether equality is important and needs talking about, and the response was that “This has nothing to do with women. It is about journalistic integrity,” and that sexism in gaming “isn’t really an issue”. From the evidence I’ve seen, I disagree. Of course, not all GamerGate supporters are sending rape and death threats. But GamerGate supporters do seem to have a women problem. Gender seems to be more of an issue than ethics here. Where’s all this gender stuff coming from?

GamerGaters feel that (feminist) critics are vilifying gamers

Lots of the complaints against Sarkeesian or ‘Social Justice Warriors’ are based on the idea that gamers are being vilified and victimized because some material in games is being called out and analysed as sexist:

  • e.g. “culture warriors, who thrive in an atmosphere of fear and moral condemnation … the entire gaming community is attacked as a pack of bigoted savages corrupted by gaming tropes”
  • e.g. “#GamerGate supporters are constantly being told they’re horrible, misogynistic, gross nerds who just want to harass women, so they’re perpetually on the defensive.”
  • (Around 22:40 in this Huffington Post Live video) “We just want to enjoy games without being told we’re horrendous people for doing so.”

I don’t think that this stance stands up to scrutiny. Anita Sarkeesian is not calling gamers terrible people in her criticism. (NB that criticism in this context means “intellectual analysis” not “saying something is rubbish”.) At the start of each of her Tropes vs Women in Games videos, she says that “It’s both possible, and even necessary, to simultaneously enjoy media, while also being critical of its more problematic or pernicious aspects.” Sarkeesian’s criticism is calm, evidenced, and highlights some games as positive counter-examples:

Similarly, Rock, Paper, Shotgun, a news site that has been attacked by GamerGate, responded with an assertion of values saying that “It is possible to criticise games and gamers, while at the same time being a gamer, loving games, loving gamers.” This is a really important point.

There is a real sense of gendered grievance in GamerGate, so let’s explore it a little more.

GamerGate is a reaction against feminism, ‘social justice warriors’ and the broadening of what is meant by games and gamers

GamerGate opposes a perceived intrusion of a liberal/left-wing social agenda into gaming and games criticism. In the words of some male developers surveyed by The Escapist:

  • Daniel Vávra: “The root cause of Gamer Gate [is that] people had enough of those hypocrites that started to inject their ideology everywhere… accusing millions of people of misogyny…”
  • Xbro: “I’ve personally been against the phenomena of ‘social justice warriors’, as well as the path certain gaming publications have taken in recent years (more and more discussions on morals, feminism, misogyny and other non-game related issues, and less talk about actual games and the industry).”

In the Huffington Post Live interview (about 9:40) one participants argues that it isn’t appropriate for a game to be criticised for being sexist in a review, that reviews should be objective and about mechanics.

I disagree here. Criticism should be, and always has been, about more than mechanics. Criticism is also about theme, dialogue, writing, art direction and pace. And if a game is sexist or racist or homophobic, I’d want to know about it because that’s a factor in how much I’d enjoy the game – and in some cases whether I’d even play the game at all.

I support criticism of games, and greater diversity and more inclusiveness in games, because it’s good and because it makes games better

I’m pleased to see games increasingly subject to intellectual and cultural analysis – it’s a sign of the medium becoming more respected. I’m also pleased to see games becoming more diverse in their subject matter, themes and mechanics, and in games being produced by a wider range of people for a wider range of purposes. The demographic of gamers has shifted so that it’s no longer male-dominated. The majority of gamers are now women (see research by the Internet Advertising Bureau and by SuperData Research).

No one’s going to stop people playing Call of Duty – particularly not as it remains massively profitable – but maybe we can consume other games that are totally different. As games become more diverse, we can play through a broader range of exciting, interesting, challenging stories and challenges. I’m pleased to see journalists respond to this change.

Rock, Paper, Shotgun stated that “We’d love gamers to mean ‘everybody’. It can mean everybody, if we let it.” I hope and expect to see games continue to develop and grow as a unique and awesome artistic medium. But to fully realise this, we need to let everyone speak, think and create. If we want gaming to flourish, let’s take ethics seriously and build an open, critical, loving community.

How to view the number of unique visitors to a page, or set of pages, in Google Analytics

You can see the number of unique visitors to your site as a whole in Google Analytics’ Audience report. But if you go to Behavior > Content Drilldown, to view a particular page of set of pages, you can only see “unique page views”.

Unique page views is a different measure to unique visitors. What a nuisance.

So the solution is to create an advanced segment in the Audience > Overview report, which includes the pages you want. (Inspiration taken from this comment by Rendy.S)

Let’s say I’m interested in the information section of the website. All of this content lives inside the /information-support/ section.

So If I wanted to create a segment to look at unique visitors to this part of the site, I’d use a regular expression like this: /information-support.*
The . and the * at the end combine to mean “plus any number of any other characters”. i.e. this will match any URL that starts

regular expression example for a site section segment in google analytics

So if you ever want to view the number of unique visitors to a particular page or pages in your website, create a custom segment and user regular expressions to target the page(s) you’re interested in.

Edward Said: Orientalism – summary

Ideas created and presented in an academic context are often brilliant but hard to unwrap and digest. I’ve attempted to pull out some quotations from Edward Said’s Orientalism that I hope will help summarise some of its key points. (Page references in square brackets are from Edward W. Said, Orientalism, Penguin Classics, 2003)

What is Orientalism?

Orientalism is a body and tradition of Western representations of the Orient, created in the context of Western political dominance over the Orient, which understand and master the inferior, inherently opposed Orient, and which bear more relationship to each other as a discourse than to the real, diverse, experiences of people who live in the Middle East.

“from 1815 to 1914 European direct colonial dominion expanded from about 35 percent of the Earth’s surface to about 85 percent of it.” [41]

“Orientalism is fundamentally a political doctrine willed over the Orient because the Orient was weaker than the West, which elided the Orient’s difference with its weakness.” [204]

“It is rather a distribution of geopolitical awareness into aesthetic, scholarly, economic, sociological, historical and philological texts;” [12]

“so far as the West was concerned during the nineteenth and twentieth centuries, an assumption had been made that the Orient and everything in it was, if not patently inferior to, then in need of corrective study by the West. The Orient was viewed as if framed by the classroom, the criminal court, the prison, the illustrated manual. Orientalism, then, is knowledge of the Orient that places things Oriental in class, court, prison, or manual for scrutiny, judgment, discipline, or governing.”[40-41]

“Orientalism as a Western style for dominating, restructuring, and having authority over the Orient.” [3]

The division between monolithic West and Orient is man-made

“such … geographical sectors as “Orient” and “Occident” are man-made. Therefore as much as the West itself, the Orient is an idea that has a history and a tradition of thought, imagery and vocabulary that have given it reality and presence in and for the West.” [5]

“neither the term Orient nor the concept of the West has any ontological stability; each is made up of human effort, partly affirmation, partly identification of the Other.” [xii]

“The geographic boundaries accompany the social, ethnic, and cultural ones in expected ways. Yet often the sense in which someone feels himself to be not-foreign is based on a very unrigorous idea of what is “out there,” beyond one’s own territory. All kinds of suppositions, associations, and fictions appear to crowd the unfamiliar space outside one’s own.” [54]
“…We need not decide here whether this kind of imaginative knowledge infuses history and geography, or whether in some way it overrides them. Let us just say for the time being that it is there as something more than what appears to be merely positive knowledge.”[55]

Debates about identity are important. Identities create outsiders and enemies

“Debates today about “Frenchness” and “Englishness” in France and Britain respectively, or about Islam in countries such as Egypt and Pakistan, are part of that same interpretive process which involves the identities of different “others,” whether they be outsiders and refugees, or apostates and infidels. It should be obvious in all cases that these processes are not mental exercises but urgent social contests involving such concrete political issues as immigration laws, the legislation of personal conduct, the constitution of orthodoxy, the legitimization of violence and/or insurrection, the character and content of education, and the direction of foreign policy, which very often has to do with the designation of official enemies. In short, the construction of identity if bound up with the disposition of power and powerlessness in each society, and is therefore anything but mere academic wool-gathering.” [332]

“A fourth dogma is that the Orient is at bottom something either to be feared (the Yellow Peril, the Mongol hordes, the brown dominions) or to be controlled (by pacification, research and development, outright occupation whenever possible.” [301]

People are more diverse than this binary

“the terrible reductive conflicts that herd people under falsely unifying rubrics like “America,” “The West” or “Islam” and invent collective identities for large numbers of individuals who are actually quite diverse, cannot remain as potent as they are, and must be opposed,” [xxii]

Antidotes and alternatives to division and Orientalism

“critical thought does not submit to state power or to commands to join in the ranks marching against one or another approved enemy. Rather than the manufactured clash of civilizations, we need to concentrate on the slow working together of cultures that overlap, borrow from each other, and live together in far more interesting ways than any abridged or inauthentic mode of understanding can allow.” [xxii]

“Since an Arab poet or novelist – and there are many – writes of his experiences, of his values, of his humanity (however strange that may be), he effectively disrupts the various patterns (images, clichés, abstractions) by which the Orient is represented.”[291]

Be aware of differences; avoid sweeping groupings; look at mingling and exchange between groups. Look at individuals and their expressions of their own feelings and thoughts. Look at self-expression, art and literature.

Recommended works:
Amiel Alcalay’s Beyond Arabs and Jews: Remaking Levantine Culture
Paul Gilroy’s The Black Atlantic: Modernity and Double Donsciousness
Moira Ferguson’s Subject to Others: British Women Writers and Colonial Slavery, 1670-1834. [353]
“In these works, domains once believed to have been exclusive to one people, gender, race or class are re-examined and shown to have involved others. Long represented as a battleground for Arabs and Jews, the Levant emerges in Alcalay’s book as a Mediterranean culture common to both peoples; according to Gilroy a similar process alters, indeed doubles, our perception of the Atlantic Ocean, previously thought of as principally a European passage. And in re-examining the adversarial relationship between English slave-owners and African slaves, Ferguson allows a more complex dividing white female from white male to stand our, with new demotions and dislocations appearing as a result in Africa.” [353-4]

“My aim … was not so much to dissipate difference itself … but to challenge the notion that difference implies hostility, a frozen reified set of opposed essences, and a whole adversarial knowledge built out of those things.” [352]

What I learnt from Coursera’s Operations Management course

Recently I completed Coursera’s Introduction to Operations Management course. The course was made up of 5 units.

Course outline

  1. Process analysis
    Measuring the flow of units through a production process; Little’s Law; inventory turns; inventory buffering: make to stock (McDonald’s) or make to order (Subway); working out bottlenecks when there are different types of flow units, processes with attrition loss; reasons for inventory.
  2. Productivity
    Lean operations and waste reduction; the seven sources of waste; KPI trees and sensitivity analysis; overal equipment effectiveness framework (OEE); reducing idle time through line balancing and standardising processes; labour, material and capital productivity; return on invested capital (ROIC) trees;
  3. Variety
    Motives for variety; batch processes and setup time; working out a good batch size; Single-Minute Exchange of Die (SMED); benefits of partial flexibility; delayed differentiation (via product design) to reduce costs of variety;
  4. Responsiveness
    Reasons for waiting: insufficient capacity and variability of arrival times and/or processing times; coefficient of variation of demand and processing time; how to compute the averate waiting time; measuring inventory over the course of a day; usefulness of pooling; strategies for prioritising work: first come first served aka first in first out, versus sequencing, shortest processing time rule; problems with appointment systems; efficiency gains are often about process redesign rather than just optimising/balancing: value stream mapping aka process mapping aka service blueprints: Vyes Pigneur’s 7 ideas for redesigning processes; waiting time and attrition loss (using Erlang Loss table);
  5. Quality
    Reasons for defects – performance and conformance quality; redundancy; impact of scrapping and rework on flow; buffers reduce risk of resources being starved or blocked, to keep flow rate up; in contrast: Toyota production system: reduce inventory to expose problems; Kanban – demand pull: work is authorised by demand, so you reduce the number of Kanban cards over time; six sigma: checking units produced against a specification; control charts: normal and abnormal variation; Jidoka system sacrifices flow for quality; Kaisan and Ishikawa diagram for root cause problem solving.

Short of summarising the outline above, I won’t attempt to share everything I learnt. Instead, I’ll share what was most relevant to my own practice.

The distinction between project management and process management

This course was about process management – about doing the same thing over and over. My job is incorporating more process management elements, so I took this course to improve my understanding, and so that I could begin to make more effective process improvements.

Some elements of my work are process management – communicating, planning, and running an agile sprint production cycle. I run through this process every couple of weeks.

Other elements of my work are project management – leading a large web development project; providing consultancy on a project; overseeing the scoping, planning and creation of a new element of functionality or a user experience improvement: each discrete piece of development work is unique. So sometimes it’s useful for me to think in terms of projects, and other times it helps to think more generically and look at underlying processes.

Little’s Law

In any process, the average inventory (number of units in the process) = the average flow rate x the average flow time (the time it takes a flow unit to go from the start to the end of the process)

Key implication: if the flow rate is constant, reducing inventory will reduce flow time, allowing work to be completed more quickly.

See more about Little’s Law.

The seven sources of waste (Taiichi Ohno)

  1. Overproduction – to produce sooner, or in greater capacities than demanded. These goods need to be stored; their production slows the rate with which you turn your inventory; they could become obsolete or be stolen.
    The solution: match supply with demand.
  2. Transportation – unnecessary movement of people or parts _between_ processes.
    The solution: relocate processes, then introduce standard sequences for transportation.
  3. Rework – repetition or correction of a process.
    The solution: do it right the first time. Find out the reason for the quality problem and put a stop to it.
  4. Overprocessing – processing beyond what the customer requires.
    The solution: make sure you have guidance for what your standards are.
  5. Motion – unnecessary movement of parts or people within a process.
    The solution: create and use standard workspaces that have been created to minimise movement.
  6. Inventory – the number of flow units in the system. The biggest source of waste. Bad for inventory turns, increases customer wait time and flow time. Inventory needs to be stored, which is costly.
    The solution: improve production control system and reduce unnecessary “comfort stocks”.
  7. Waiting – underutilising people or parts while a process completes a cycle. i.e. a flow unit waiting for a resource. Often a direct result of inventory. Waiting can happen at the resource: this is idle time.
  8. Intellect – an eighth source of waste. Don’t waste workers’ abilities to help solve problems and improve processes.
  9. Increasing profitability is easier if you’re constrained by capacity than if you’re constrained by demand

    If you’re constrained by capacity, increasing the productivity of the bottleneck can help you significantly increase profits. (This is particularly the case for businesses with low variable costs and high fixed costs)

    If you’re constrained by demand, you’ll only be able to significantly increase profitability if you’re able to lay off workers.

    Variability increases wait times, even if resource utilisation is less than 100%

    If people arrive at regular intervals, and take a fixed length of time to process, then you can plan your processes to avoid waiting time. But real life is less predictable.
    Variability of arrival times and processing times can lead to inventory in a process, even if utilisation is 80%.
    So variability means that even if you aren’t utilising all your resources all the time, you’ll still have people or products waiting in the process.

    Two reasons for waiting: insufficient capacity; variability of arrival times and/or processing times.

    If you’re constrained by capacity, you don’t need to worry about demand variability as you already know there will be bottlenecks. If demand is the constraint, and it’s variable, then you need to think about it, as it will cause waiting times.

    Partial flexibility is usually the best way to deal with variety

    If there is variability in demand, you need to accommodate it. Total flexibility is expensive, and usually not needed.

    Eg. It’s sensible to hire developers who have skills in two areas, so you have flexibility, but don’t have to pay the costs of a developer skilled at everything.
    For each area of your work, hire at least two people with those skills.

    “The way we frame a problem determines the types of solutions we come up with.”

    A surprisingly philosophical insight. Often it pays to be more creative than just doing queuing analysis and line balancing.
    Question your processes at a strategic level – don’t just think tactically and inside the box.

    Value stream mapping is a tool to help you focus your process on valuable activities

    Value stream mapping, aka process mapping, aka service blueprints – map out the steps the customer has to go through, then divide them into ones that add value and ones that don’t, or which are waiting time.

    Yves Pigneur has a framework for this: Customer actions; onstage actions, backstage actions; support processes.

    7 ideas for redesigning processes

    1. Move work off the stage.
      E.g. online airport check-in.
    2. Reduce customer actions / rely on support processes.
      E.g. rather than requiring customers to fill in all their medical details each time they come to visit, you could have a database to store them.
    3. Instead of optimising the capacity of a step, see if you can remove it altogether if it isn’t really needed.
      E.g. Hertz Gold removed the airport check-in step as it provided no value.
    4. Avoid fragmentation of work due to specialization / fragmentation of roles.
      E.g. in a bank, it’s annoying to have to fill in different forms for different people, rather than just doing everything at once.
    5. If customers are likely to leave the process due to long wait times, move the waiting time to later in the process if you can.
      E.g. Starbucks making you pay first, then wait for your coffee.
    6. Have the waiting occur out of a line.
      E.g. restaurants in malls using buzzers to let people know that their food or table is ready, rather than having them wait in a line.
      E.g. appointments to see a doctor.
    7. Communicate the wait time to the customer – set expectations.
      E.g. theme parks.

    How much do defects cost? It depends on where they are detected.

    If defects are detected before the bottleneck, the cost is driven by the input prices.
    If defects are detected after the bottleneck, the cost is the opportunity cost of the lost sale.

    The step at which the defect happens isn’t important – what’s important is the step at which it is detected.

    Therefore it’s very important to test flow units as much as you can before you put them into the bottleneck.

    For me, this means that we want to catch problems in the specification stage if possible. Minimise potential problems by making sure that the work is sensible, and that the requirements are clearly articulated.

    Kaisan and Ishikawa are tools for root cause problem solving

    Kaisan equips front line workers to identify and solve problems.

    Ishikawa diagram – structured brainstorm. Shaped like a fish bone. Try to identify root causes. Asking ‘why’ 5 times helps.

    Once you’ve done this exercise, go out and measure instances of the identified defects.
    Plot these on a pareto chart.
    See which defect is most frequent and focus on that first. Generally the pareto principle applies: 80% of the defects are caused by 20% of the root causes.

    These methods are recommended because they oscillate between thought and reality, gaining the benefits of both:

    Reality: Jidoka – the process is triggered by real-world defects.
    Thinking: Ishikawa diagram, to think about what might be causing the problem.
    Reality: Pareto chart – collects data to see which causes are most frequent.
    Thinking: Think up alternative solutions.
    Reality: Experiment with the solution you choose.

    Actions I will take as a result of this course

    • Map out the production process so everyone knows what steps are required. If people don’t understand all the steps in a process, they might have unrealistic expectations of waiting times.
    • Create formal processes for new jobs – development/UX work and bugs. This will reduce variability of inputs and reduce risk of defects by improving the quality of inputs/briefs.
    • Formalise waiting time processes. Draw up (collaboratively?) and obtain organisational agreement for a set of organisational priorities for bugs and for new work. This will mean that all prioritisation decisions are made according to a clear set of standards. E.g. number of users affected, financial implications, strategic priorities. Currently I don’t do first-come-first-served, but rather prioritise according to business need and urgency, but this does require me to be wise like Solomon. Better to have some commandments to live by, and a supreme court to interpret them.
    • Reduce the number of units being processed at any one time. Little’s Law states that this will result in reduced waiting times. This might be a hard sell, but the truth is that we are already constrained by how many hours of development we have in each two week cycle.
    • Identify and reduce the sources of waste in my work. eg transporation – reduce transportation costs in communication with agency and stakeholders (make it clearer), reduce movement of work around internal stakeholders (currently I report to internal stakeholders outside of the tools I use for day-to-day project management, which adds costs and the risk of misinformation.)
    • Harness worker intellect through more regular review cycles. Constitute regular reviews of processes with all people involved in them, to see how they feel they are going, and what could be improved.
    • See if processess could be redesigned to be more efficient. Could steps be removed or automated?
    • Conduct value stream mapping – figure out which steps add value and which ones don’t. Use Pigneur’s framework for process redesign to improve these processes.