When ecommerce goes wrong
Rob Smith

Reading time: < 1 minute

Haven’t posted for a while but will pick up the pace again soon. In the meantime, here’s a great shot I got off a pet supplies website that is largely quite well put together but this promotional message made me chuckle:

0% saving!

Related Posts


Do I need usability testing?
Rob Smith

Reading time: 1 – 2 minutes

We get asked this question a lot, especially when we suggest that a client needs to run some usability testing to make sure the solution we’re proposing works for their target market.

The reason we get the question

“I’m paying you to design and build the right website for my target market. You’ve done the research into the market and you’re also the professionals. Surely if you have done your job right, then it will work great and doesn’t need usability testing?”

It’s actually a pretty fair argument when you think about it. It’s not quite accurate though. No one ever truly knows everything about how a particular target market works or behaves, especially when introducing something new to them.

The answer

Recently, I heard an analogy on why you need usability testing. I can’t remember who exactly said this although I’m certain it was while I was at SxSWi this year. It’s this:

It’s the story of the tailored suit. You can select the fabric, take the measurements, select the style, and create the suit. But it hardly ever fits first time. That’s why you go for a fitting and the master suit maker nips and tucks around you while you’re wearing the ‘nearly there’ suit. Usability testing is that fitting session. It helps us before those nips and tucks to make the experience perfect.

It was the best explanation of why you need usability testing I’ve heard for a while.

Related Posts


Video: Me at a roundtable at UKFast
Rob Smith

Reading time: < 1 minute

A few weeks ago I attended a UKFast roundtable talking about the marketing mix and how it’s changing especially with more and more digital marketing channels opening up. Check out the video below to see me rambling on with a few other clever Manchester people.

Related Posts


How to structure a digital agency
Rob Smith

Reading time: 3 – 4 minutes

It’s clear from some of the traffic coming to this blog that there seem to be quite a few people wondering how to structure a digital agency – what the options are and therefore the pros and the cons. I thought I’d write a quick post from my experience of running an (albeit small) digital agency and how we work, and other models I’ve also seen around.

How does Blueleaf Digital work?

Well as a small, collaborative team we have the standard parts: design, development, project management, admin.

We work together, close to each other, and so there’s constant communication between all departments that is face to face. This works really well for us because we can get together and sort problems or challenges out quickly and easily by calling a quick 5 minute get together.

Generally websites go from sales to planning and project management to design to development to testing to launch. They go back and forth between project management, design and development. Our process is shown below. We’re always tweaking and changing it but this is our latest iteration (click to enlarge):

Blueleaf Digital website process

This is really just a website process however, it’s not the same process we use for all types of project. Having said that, a lot of variables remain the same.

For a small agency, this is generally the structure, a highly collaborative single team that works together an is pretty agile.

What happens when you get bigger?

As an agency gets bigger, I’ve seen three different routes to structure the agency, and there are some pros and cons of each method.

1) Big teams approach

You can keep growing each team, design team, development team, copy team, creative team, and so on. Each team just keeps getting bigger, and there is a lead in each team that distributes the work the team needs to do.

Pros: Easy to manage and build, great depth of knowledge in one place
Cons: Builds a our team is better than yours / competition by discipline

2) Mini discrete teams

Build lots of different teams that work together on a variety of projects, a PM, some designers, some developers. Small and agile, this method mirrors a lot of smaller agencies, in effect creating a small agency feel within a big organisation.

Pros: Small and agile, good teamwork
Cons: Not a great depth of knowledge per team, team does things same way each time

3) Mini teams with swapping

This is a variation of the 2, with the team being formed on a per project basis. This means that each member of the organisation will be on several teams doing several projects at once.

Pros: Small and agile team, broad experience due to working in many teams
Cons: A lot of resource management required to know where resource gaps are

As we grow, we’ll be looking at option 2 primarily, and maybe give 3 a go for a while if we get much bigger. It;s important that our structure:

  • Motivates our staff
  • Keeps feeding them different work
  • Keeps feeding different peoples viewpoints and experience
  • Keeps overheads and administration costs to a minimum

Comments welcome on how your agency may be structured, or how you feel it should be!

Related Posts


Don’t just hunt, farm
Rob Smith

Reading time: 5 – 8 minutes

The Internet and Ecommerce have really taken off, especially in the last five years. By all accounts, this is going to continue. Recently though, there’s been much speculation about Internet growth rates and whether they and the revenues derived from them are really going to continue growing at such an astronomical rate.

The general consensus is no.

Verdict consultancy recently released a report (November 2009) that states two trends that we need to be highly aware of which I hav

  1. Growth will become more difficult: Through a combination of factors such as a slowdown in broadband adoption and increasing maturity, there will be less new customers appearing online than in previous years.
  2. Purchase process: Customers will visit more sites than ever before making their purchase, plus there will be more retailers to choose from.

Combining these two factors will mean that the traditional acquisition model used to grow by a lot of current retailers will not be enough to meet their new targets. We can no longer just be on the hunt.

Most organisations are not farming

So what else can we do? Well a lot of retailers have started to farm their existing databases (customers) to try and get them to purchase more from them. It’s far cheaper after all to convert a current customer again than bring on board an entirely new one.

This of course makes good sense. Unfortunately, the general approach is quite flawed and does not take advantage of the much greater potential that farming can provide. Very often the approach is centered around bulk email sending to a customer database with several offers on the email. We have no idea whether the customer will like the offer presented to them. It’s a simple numbers game of if we throw enough email at the wall, some of it will convert to orders.

Generally this has worked quite well for retailers and has therefore earned email marketing the reputation is deserves of being a low cost delivery mechanism with high (if gradually eroding) returns.

This is not farming and it’s not hunting. It’s looking at your land and using a giant cannon to fire food/nutrients at your land. You have no idea which land grows what but let’s just do it anyway.

How to farm

So, having said all of this, what is farming? Farming is caring for your land (customers) and giving them the right food and nutrients to create the right results. So let’s drop the metaphor – it’s about highly targeted, well delivered messages based on not only them as a person but their behaviour.

Clean

To create an effective farming strategy, we first of all need clean data. Your database is probably not clean. In fact, it’s probably a mess. It’s been added to over time, not used very much or maintained, and could be in multiple places and with different types of data. We need to:

  • Bring all of the different data sources together
  • De duplicate this information to have one single set of data per person
  • In the case of email data, check that all the domains exist and can accept email and all email addresses are in the correct format

Now our data is quite clean, but we still don’t know whether the actual email addresses exist as this is a problematic area. Therefore our first send will clear out a lot of dead email addresses. If you have already been sending regular emails, al lot of this will already be done.

Segment

This is the crucial stage. Currently we have a mass of data that still can’t be used for farming. So we need to create smaller groups of data based on their preferences and behaviour. We’ll take an ecommerce site as an example. In this case you should have the following information:

  • Who they are (contact details, name etc)
  • Where they are (address information)
  • What they have bought (purchase history)
  • What they have returned (surprisingly important)
  • When they buy (time of day, week, month, year)
  • What they have responded to before (email activity history)
  • Where they came from (original source of customer)

The list above is by no means extensive and you may not have access to all of the data, but it will serve well for our example.

Based on the data above, we need to find the pockets of customers where correlations can be observed. For instance are a high % of your customers in the South East? Do people who are in the north only buy certain products from you and not others? Do people generally seem to buy in the evening, weekends or is there no pattern? All of these kind of questions will help you identify pockets of highly concentrated data that share behaviour.

Target

Finally the action part. Once we have identified some great pockets of data we need to use their behaviour to send them the right message, at the right time.

Let’s say that we have 1,200 customers who seem to buy at the weekend, are located in the West Midlands and really like the electronics section of your website. That’s pretty specific I know, and sometimes it’s not always possible to get that specific but we can always try. We should segment as deeply as there is a strong correlation of data. 1,200 people exhibiting the same behaviour is significant, 3 people is not!

So we have the data, what can we send? Well the idea would be to send an email, possibly on a Friday afternoon or a Saturday morning or even a Sunday morning (all at points in the weekend) with an electronics specific offer. If there’s something massive happening in the West Midlands that weekend, like strong gales or heavy rain, the campaign could focus on that.

If you were one of those 1,200 customers who received that email, are you more likely to respond? According to your behaviour you will!

Conclusion

Yes, it’s a lot of work, and it’s worth the effort. The returns that can be gathered from highly specific campaigns sent to very segmented data are much more than the returns of bulk mail shots scattered out. The hard work is also in the initial clean and segmentation. Following this process you should tweak change and retest or segments every so often (6 months or so) but generally it’s just coming up with the right messages to send to your segments.

So please, start farming as well as hunting. If I get another email inviting me to a women’s only networking event, I think I’ll scream.

Related Posts


5 important ecommerce themes, 2010
Rob Smith

Reading time: < 1 minute

Just published is my new article over at Think Vitamin, Carsonified’s blog. It’s about surprisingly 5 important themes for this year in ecommerce. These are generally themes to look at if you have all of the basics like well converting pages, easy checkout, etc all sorted. Check it out:

http://carsonified.com/blog/business/5-important-e-commerce-themes-for-2010

Related Posts


You can make data lie to you
Rob Smith

Reading time: 2 – 3 minutes

We are coming into a digital / data age by all accounts, and have been for a while. Everything is based on data, data data and making decisions based on that data, especially for ecommerce sites. Really though, do we know what we’re looking at? Do we understand the numbers? The differences in conversion rates, visitor sources and the like?

The truth is that data has two major caveats that everyone needs to be aware of and needs to put their statistical hat on to understand:

Data can tell any story

Depending the on the way you cut your data, the way you include or exclude certain groups and the metrics you use, data can tell wildly varying stories to people. The data, of course, is always the same (hopefully). One pool of data, and we attach meaning to that data. It’s the meaning we attach that can differ so much!

If we go into data wanting to draw a particular conclusion, then we can find a way to do it! Be careful.

Data needs to be statistically relevant

The good old AB test. A dangerous tool at times. We’re comparing two different landing pages. One converts the visitor 10% better than the other one. We’re all happy with the work done. Is this number real? ‘What do you mean is it real?’ I hear you cry – how can numbers lie? They can’t. But they can hide the whole truth.

First of all, sample size. Did you test enough visits and conversions. It ran over 100 visits? Not enough! To get statistical relevance we need a large sample size, over a decent time period.

Secondly, if you try an AA test, comparing exactly the same creative to itself, in theory they should perform exactly the same. In practice, this does not always happen. In fact, over at Get Elastic they performed an AA test and one version outperformed the other by 4.97%! Shocking. So that 10% difference could of been 5%, or 15% – who knows!

If you perform tests make sure you have statistically relevant numbers, and that the results you get show a very clear difference between the two. If you can, perform an AA test also.

Related Posts


Agency structure vs Client structure
Rob Smith

Reading time: 5 – 8 minutes

This week I attended one of the Marketing Industry Network events in Manchester which discussed the future of the independent digital agency. The panel including agencies like Cool Pink, Code Computerlove and Chapter8 and come question time I asked whether the change within the industry is been driven by clients or agencies. It was interesting discussion and inspired this post which will explore how the structure of the two sides of the agency/client relationship seem to affect the working relationship, projects undertaken, and results gathered.

We’ll cover this is this order:

  • Agency structure
  • Client structure
  • How this affects projects and results
  • The future

Agency structure

Agencies can obviously be structured in many different ways and there are probably more ways than I could ever think of. Having said this however we can normally say that there is what is seen as a standard agency model driven by departments.

One department does the creative, one the technical build, one the testing / QA, one media planning and so on. Projects are passed from department to department to be worked on and generally go back and forth. The departments don’t always communicate that much with each other. They are lead by the account director or executive who’s the communication path to the client. This model is a pretty natural output of a growing company in it’s attempt to cope with scale.

I will call the above the silo approach. Team’s are quite separate and not all that collaborative.

The other way of working is focused much more on collaboration. Whether it’s a lot of smaller teams with someone from each discipline in each, or, in the case of a smaller company, one giant collaborative team.

Client structure

A similar situation exists within clients as well. A spectrum between collaborative and silo based organisations. The difference with clients is also where the decision makers lie for different projects. Where they lie can make a dramatic difference to what projects are prioritised and why, and how the relationship can be affected.

First lets discuss structure. In my experience there are a lot of different structure possibilities within organisations and a lot of this is dependent on how they have grown. Here’s a couple that we have come across in charge:

  • Business managers / Business development
  • Direct Business (If the company has a retail arm)
  • Managing director / CEO
  • Marketing
  • Ecommerce
  • IT

It makes such a huge difference who is responsible for a digital/ecommerce project as to the results which we’ll discuss later. The stereotypes of the above departments / decision makers can be very true. IT are generally more interested in how much work it means for them and security (rightly so), business development and CEOs are very much about the bottom line and marketing tend to be more aesthetically orientated.

The key with how the client is structured however is less who is the point of contact, and more about the level of collaboration within the organisation. If each part hardly ever talks, it means something very different to if they have regular catch up and joint planning.

How this affects projects and results

If you’ve worked at any agency for a period of time, you know some projects go better than others, some seem simpler, even hen the project is more complex. Often the less complex the project, in a paradoxical way, can be harder.

I believe that the alignment of client and agency structure can play a big part in how well they work together. You could argue that a good account handler should protect the agency and client from any differences that can cause friction. To a certain extent that’s true. I do believe though, that the underlying structure and therefore culture of the organisations will always come out.

Combo One – Collaborative client, Silo agency

If a collaborative client tries to work well with a silo’d agency, they will become frustrated that things cannot be more collaborative. They will crave to have good, productive meetings with all of their major players, and all of the agency’s major players.

Combo Two – Silo client, Collaborative agency

If a silo’d client tries to work with a  collaborative agency, the agency will become frustrated by the fact their ideas and forward thinking fall often on deaf hears depending on the contact within the client. Ideas and work are constantly filtered through their eyes resulting in their view dominating. When work is finally presented often other organisation stakeholders are disappointed as it’s so focused from one area. More work for the agency.

Combo Three – Silo client, Silo agency

If a silo’d agency works with silo’d client, things go very, very slowly. Communication is drawn out, Chinese whispers are prevalent. Work takes a long time to get done and there’s a lot of back and forth from agency to client and inter department within both client an agency. Even after a drawn out process, the work produced will end up not really pleasing anyone either side unless there’s on dominant personality as the client contact.

Combo Four – Collaborative client, Collaborative agency

If a collaborative agency works with a collaborative client, projects can go exceptionally well (in our experience). The stakeholders get round the table and talk out what they want from the project. Everyone’s views get aired and they feel like they have been heard. The agency then works together on the project, discussing the possibilities for the project. Again, all parts get heard and discussed. Projects move faster and results are generally better. There are a couple of caveats. Trying to get all the stakeholders together can be difficult and therefore slow things down waiting for a meeting. Sometimes both clients and agencies can get too many people involved with too many view, and a project can lose direction and clarity. You can never please everyone.

Conclusion and the future

So what does this mean? We can’t all magic the right clients or change our current clients, organisational change takes time. I think one of things all agencies need to realise is that one of the reasons digital agencies seem to be doing well is that their attitude to collaboration and change. Digital moves so fast as an industry, that the agency has to be collaborative and good at change. This helps them with clients as well, to learn and adapt.

In my opinion, all agencies must work hard to be as collaborative as possible and encourage clients to do the same.

Related Posts


Mobile marketing on Wise Marketer
Rob Smith

Reading time: < 1 minute

My article on why context is so important to mobile marketing, and how that differs to normal marketing has been posted over at Wise Marketer. You’ll need a free login to read the article (well worth the effort) and I would thoroughly recommend registering for their daily update – so much great content comes from these guys (if I do say so myself):

http://bit.ly/bcKoDb

Related Posts


Do you just collect data? Or do you analyse?
Rob Smith

Reading time: 2 – 2 minutes

Doesn't matter how many screens you haveMost companies are collecting data for no reason. Far too often we’re coming across companies that have some form of analytics but no idea why. They have had it put in (mainly because it’s free) and someone has normally shown them how to log in and told them what some of the numbers mean. They have never logged in again, let alone actually done any analysis of the data.

Data is useless without meaning. Meaning is useless without action. Companies need to be analysing their website statistics, finding out why certain trends are appearing or not appearing, and take action to change what they are seeing. Only then are you truly using web analytics, and not just being a data collector for no reason.

Here are some things you might want to know about your visitors:

  • Where are they coming from? Which of these sources is the most profitable? What’s the trend of these visitor sources? Which are declining or increasing over time?
  • What is my bounce rate? Which content gives the worst bounce rates (and therefore should be fixed)
  • What content are they looking at on my site? Which pieces of content are the most popular? Which produce the most exits from the site?
  • What keywords are people using to visit my site? Are they all brand led? If so what can I do to encourage people to visit my site that haven’t already heard of me?

There’s a ton of questions you can ask yourself. So why not start asking yourself right away. Stop collecting, start analysing.

Technorati: 3Z58EZ26P9TB

Related Posts