July 13, 2009

But is it viral?

Wikipedia defines viral marketing as:

“marketing techniques that use pre-existing social networks to produce exponential increases in brand awareness, through self-replicating viral processes, analogous to the spread of a computer virus. It can often be word-of-mouth delivered and enhanced online; it can harness the network effect of the Internet and can be very useful in reaching a large number of people rapidly.”

Or how about a simpler version:

“Compelling enough to tell someone else about”

Increasingly we are seeing companies trying to “make it viral” for their online and offline marketing campaigns. A lot of these focus on the mechanics of viral, but we think this misses the point entirely. To be truly viral, the idea has to be compelling enough to encourage people to spread the word without being prompted.

The contenders

The approach: RealEstate

RealEstate’s concept is to donate space on a big billboard that lots of people drive past. The cause is a good one; promoting awareness of foster homes (hence “do something big for kids”).

The proposition is that they have hired this billboard. If 100,000 people hit the Do Something Big page on their site by Nov 14, they will donate the billboard space for the remaining rental period. Get your friends to come and click on the page to get to the target earlier.

Real Estate

The approach: Threadless

Threadless’s concept is nothing new : give something away to reward ongoing patronage. If you’re a member, you see the Loves page, where all the current promotions are. The latest of which is 550 free tshirts. For every tshirt you buy, you get an extra entry into the draw. If you take a photo of yourself wearing one of the shirts, you can upload it to the site and earn money towards your next purchase. And if that wasn’t enough, if you refer a friend to the site, and they buy a tshirt, you get even more money off your next shirt.

Threadless

What worked, what didn’t: RealEstate

1 – Conceptual issues

The idea itself is a nice simple one that ties in well with the brand- helping find homes for kids in need. Perhaps it all started as an idea about running a billboard, and then someone suggested tacking on a viral web strategy thingamejig. You can almost imagine how little “improvements” were added in:

No, let’s not just give it away- let’s advertise that we’re going to give it away, and then give it away” then, “Hey, we need to also drive web traffic- lets make people come to our site first!

Surely it would have been simpler, cheaper and more effective to just donate the banner?

2 – That url
  • too long to remember
  • people don’t “get” underscores. They look too much like an underline, and are a pretty unusual character for a non-webdeveloper to ever need to use.
  • the banner was positioned in a way that made the underscores hard to read when the banner is viewed from below
3 – And the point of this is?
  • No mention of benefactor in online and offline banner, so most people wouldn’t realise that it wasn’t actually an ad for RealEstate
  • Gets less attractive over time- they only have the billboard for 4 weeks. They reach the target, and then flip it over. This makes it most attractive right at the start, but the attraction decays over time. Week 1 and 2 it might be worth the effort, but come week 3- is it really worth the effort to get your friends to visit if it means a week of donation? All the while RealEstate has been advertising themselves for the other 3 exclusively – it just doesn’t seem right.
4 – You get there, and…
  • And then what? The experience was a bit underwhelming – it would have been good to at least find out some information about the problem that DOCs is solving, or other ways that you can help. Or other initiatives that RealEstate is involved with. I’m sure they are very fine people and support some great causes, but this campaign doesn’t really communicate this. Failing to take people to the post-viral landing page was a big mistake.
  • There is a send to a friend thing (albeit simple) – but just by having this there doesn’t endow the page with magic viral qualities. It has to be compelling enough for someone to want to tell people about.
5 – Oh- and the counter doesn’t work
  • If the key “thing” is hits on the page, then its important to make sure that it works well. When we visited in the early days of the campaign, the website counter didn’t work- a small but significant problem that undermines trust.
  • A better approach might have been to explicitly acknowledge that YOU have just made a difference, Mr IP number 123.123.123.123 from Surry Hills, NSW, Australia. And tick it over in real time so that you can see other people hitting the page after you.

As of today, they failed to reach their target by the suggested deadline. To the credit of RealEstate, they have gone ahead with the new outside campaign regardless. And it looks great. There’s even a shiny new URL – homesforkids.com.au. Why did they not just do this in the first place?

The original Do Something Big page had recieved some 80,000 hits by Nov 14 (according to its counter), but even if every one of those was a unique visitor, they would have had NO WAY of getting to the new Home For Kids site, unless they happen to see the new outside billboard. So any good will that was generated by the first half of the campaign has now gone to waste.

What worked, what didn’t: Threadless

For Threadless, the proof is in the pudding. As highlighed in a previous article on 37Signals, this small company has in the past year sold more than US $6 million in stock, and plan to triple that in the next year.

People are talking/blogging about it

So popular is the site that there are fan sites dedicated to news and competitions from the website, forums set up specifically to talk about content on the site, and all of their competitions are made with the specific desire to increase brand awareness.

It’s easy to understand

The campaign is simple to understand – buy stuff and we’ll reward you. Tell your friends and they can win stuff too. Tell us about it in a photo and we’ll reward you.

Real impact

What’s more, all the campaigns they run are accountable, because they are based on real users who really want to engage with the content (and really want what Threadless have to offer them in return). And other companies love Threadless – which itself spawns much of the giveaway content on the site.

But is it viral?

After all this analysis, we are back to the initial question. One way to measure this would be to look at the final numbers for each of the respective campaigns.

Another much more subjective way is to apply the “Pub Logic” test (thanks Justin): Imagine you’re at a noisy pub with a friend. Is it simple enough to explain it quickly between beers? Secondly, is it worthwhile telling them about?

Realestate.com.au – nup

Why?

  1. All external advertising (outdoor banner and web banners) came across as a corporate advertising themselves rather than a worthy cause
  2. You couldn’t do anything about it once you saw the banner – apart from try to remember a very long URL for when you’re next online
  3. The reason to tell someone else about it wasn’t particularly compelling (particularly towards the end) and hard to explain
Threadless – yep

Why?

  1. There’s a direct reward in it for me
  2. They encourage people to talk and blog about it
  3. pretty easy to understand how it works
July 13, 2009

Building a meaningful forum

One of the things that we’re constantly on about here at Red Ant is meaningful interaction. A website is all well and good if it gets a gajillion hits, but, if those hits are from bots or the human equivalent, then the only meaningful thing that you’ll have at the end is a large bill from your hosting provider.

One of the ways people have interacted with one another since the Internet was in Internappies is, of course, through forums. And, while many forums can lay claim to fostering community and interaction between users, just as many forums lay waste to that same idea, and leave behind only wasted time, wasted words and wasted bandwidth.

When Huggies (did we mention it was one of the most popular parenting sites in Australia?) asked us to rebuild their forum earlier this year, we had all of this in mind. But how’s it gone so far?

Well here are some stats: Over 12,000 posts have been made by over 1,200 individual members in one week. The longest post has over 2,800 replies and counting. It’s had ‘more than 600,000 posts since it was launched, and the company’s experts are now receiving more than 100 questions a month.’

This is all summarised neatly in an article in the Sydney Morning Herald:

‘Through the forum, Huggies is able to build a strong relationship with parents that hopefully builds loyalty to Huggies products. ’The options for communication and interaction that web 2.0 technologies present are widening rapidly.’

There are a few things that mark out the Huggies forum. The presence of experts helps to guide user-generated discussions along helpful paths – users can submit questions and get them answered on a one-to-one basis, as well as discuss related answers together. It also helps that the forum is focused around a singular topic, parenting and pregnancy, and while it might seem to be an obvious thing to say, that topic is one that the core audience is keen to discuss.

Lesson for the day, when you’re building interactivity for users: spare a thought (and hopefully a bit more) for what that interactivity will mean, both for the user, and for the client.

July 13, 2009

A tale of two bumps – Youtube vs Superbowl

Adage has published an article that looks at the current Evolution campaign for Dove, and compares the traffic resulting from a Youtube spot vs a Superbowl ad. Here are the statistics from Alexa, showing performance:

alexa graph showing site traffic on www.campaignforrealbeauty.com site

The first bump is traffic from the Superbowl ad, which cost around $3.75 million to run. The second bump is from this ad on Youtube, which ran for free:


The results are pretty compelling- the first spike jumps and then falters, but the Youtube is massive initially, and then appears to be gradually tapering. Keep in mind that this isn’t traffic watching the video itself (that is handled on the Youtube site) – this second bump is clickthroughs from people that have seen the video, and then are interested enough to click through to the site itself.

From the Adage article:

“This is a great example of where we’re not using the old playbook where we do a lot of TV advertising,” Todd Tillemans, VP-North American skin care said. He believes the strong consumer insight behind “Campaign for Real Beauty” gave the effort “viral legs” and that the particular message was “more powerful because it came from an objective source” in the form of the TV news and entertainment programs.

But don’t forget that Youtube does not = Viral – manufactured viral can cause more damage than you might expect. Just ask those hip dudes at Agency.com, who dreamt up a viral approach for a Subway account pitch but then pulled out suddenly after creating much controversy and mirth.

July 9, 2009

How we Wiki

One of the big changes that we have made over the last year or so is the way that we communicate with our clients. One of the big challenges is to avoid drowning in the tidal wave of information that comes with a typical web project. There are literally thousands of ideas, meeting notes, concepts and diagrams continually popping up, and they’re tough to keep abreast of.

To do this, we use a type of software called a Wiki. A Wiki is what you’d get if you had a virtual cocktail blender and mixed Word with a Blog and then threw in a Content Management System. And yes, Wiki like in Wikipedia.

But we use our Wiki in a slightly different way. Rather than just a receptacle for information, we use it to store meeting notes, brainstorms, wireframes, snippets of code, and even present visual designs. The aim is to try to get as much information on each project written down. By doing this, we’ve found that our Wiki has become our main interface with customers – a kind of a mixture of intranet and extranet.

The actual software we use is called Confluence, which is made by Atlassian. They did an interview with us as part of a customer case study:

“We tried a number of different approaches, and looking back they were all based on a central editing model. A wiki made sense because we wanted everyone to be able to contribute and participate. It is also closer to the way that we like to work with our customers. By allowing everyone to be able to add and reshape content, more people became involved. We moved from one person slaving away creating pages and the rest of us having to wait for them, to a situation where one person gets the ball rolling, and then other people can join in to complete the task.”

“Say, for instance, we’ve created a design and need to show it to our client. First, a designer makes a page, attaches an image, and they’re done with their part. But then I might look at it and realise that it needs a bit more explanation, or a link to a wireframe diagram to give context. One of our developers might have also mocked up how a menu works, and so they stick in a link to that. Our client might email the link around, and then add some comments on the page. This kind of collaborative workflow is one of our strengths, and it is really important for us to be able to add these various types of content easily.”

You can read more here.

June 13, 2009

Tracking stuff – part one

A while back, we started looking around for options for tracking bugs that came up in the process of developing web sites. This is a summary of what we learnt, and some of the steps that we went through.

The basic problem; you have a project where there are lots of things that you need to write down and keep track of. These need to get resolved before the project can wrap up.

Initially, we used issue tracking for tracking bugs- identifying and recording things that were broken or not working properly, and then recording when they got fixed. Looking again at it now, about a third of what we do are “bugs”, the rest would be better described as tasks – so project management rather than bug tracking.

We started off with the “low tech” approach- using an Excel sheet. Mine had a description of the issue, a unique number, and some kind of way of marking issues off. As the project progresses, these were updated and hopefully checked off. The drawback to this approach is that it’s quite manual, and needs an “owner” to update. We found this approach not particularly scalable – it worked OK sometimes, but on a rapidly changing project it became unweildy and got out of hand.

The first issue was pretty simple- version control. The excel file got mailed around, and people would start to update or annotate their version. Then these changes would need to get merged in to an update version. The first or second time, it’s not that hard; but gradually things start to get just too much hassle. Yes, I’m sure that the solution lies with some kind of Advanced Merge With Tracked Changes command, hidden deep in the bowels of Excel interface, but I could never find it.

Another problem comes with distribution- how to let everyone know about the list, what they have to do, and what other people in the team are working on. Emailing it out as a “broadcast” works, but then we ran into version issues when people started using it. Again, I’m sure there is some clever function in Excel to make nice HTML, but I always seem to end up with a tag soup mess. And then people start emailing in updates, and some poor soul then has to reintegrate these into the spreadsheet.

The next issue started off as a filtering problem- some people need a list of just their issues, while other people (like a producer) needs a list of everyone’s. Then the customer needs to get a feel for what has already been resolved vs what is left to do. But then we had one project where we needed to manage things that were being done by freelancers, as well as our internal team, and create a nice report back to our customer. The whole Excel plan started groaning at this stage. Then there are those “secret” bugs- a few things that we’ve stuffed up and needed to get rectified urgently, but we didn’t necessarily want to broadcast this to our customer and their boss.

What really sank the good ship Excel was what I call “issue blossoms” – when one seemingly simple, innocent sounding issue suddenly spawns 20. They seem to appear in the later part of a project, when you’re getting in to the meat of the project (and everyone is strapped for time anyway). You can try to have some kind of numbering system like 2.4.1, 2.4.2 etc, but you’re starting to walk the thin ice by this stage. Sub list spreadsheets are another way of handling this, but these introduce lots of other issues.

Newer Posts
Older Posts