MeasureFest 2013 > In-Depth Takeaways
When it comes to topics such as conversion rate optimisation, web analytics and marketing measurement, not many of us would believe there are specialists out there who would be willing to put on a whole day of educational insight to help us get the most out of all of these subjects.
Well luckily for us, on 17th October 2013, the brains behind BrightonSEO and the Content Marketing Show brought together the first conference of its kind designed to do just that – in the form of MeasureFest.
Myself and fellow Zazzlers were lucky enough to bag ourselves tickets, so we headed down to London for the day to bring you all of the best insights and main takeaways from the CRO, Measurement and Analytics decks of the day:
Paddy Moogan, Distilled – A Framework for a CRO Project:
For those that want to view all of Paddy’s deck, you can do so below:
- From the outset, CRO looks easy (change a bit of text/colour), but the reality is it’s not that simple.
- In fact, if you don’t know what you’re doing, you can screw things up.
- As agencies, we don’t always have the time to focus on CRO and hiring the right expertise isn’t easy, especially if you’re not known as a CRO company.
- Expertise requires live projects. You can read all the info in the world, but until you’re actually doing it, it isn’t quite the same.
- And at the end of the day, clients won’t give you work if you can’t prove expertise – it’s a vicious circle.
- But actually, you don’t need to be an expert straight away. You can start off small and can be can be good enough to be where you want to be…
- …It’s all about the ‘learning curve’. To become a true expert in any topic, research shows it takes about 10,000 hours to do so – five years in total.
- But actually to know enough, all you need to do is get to the top of the curve. Here you’ll know 95% of what you need to know and still be good enough – this takes only 20 hours, compared to five years.
- Can use a simple CRO framework to help clients well.
A Framework For CRO Projects (That Everyone Can Use)
Phase 1: Discovery
- Understanding The Business
- If you just dive into things, people don’t understand why they are doing it.
- Benchmark current performance – if you don’t do this, you can’t show the value of your work.
- Sit down with client and clearly define business goals.
- Figure out why the company exists – helps to show the decisions they make as a company.
- List Unique Selling Points over competitors – identify them.
- Map out the sales process from start to finish, as well as all touch points – including offline.
- Understanding The Customer
- It’s really important to list all fears, concerns and needs of the customers and think about what they care about when looking at a web page. Your concerns will be different to theirs.
- Identify the demographic of target customers – especially if you work internationally. Because of the different cultures, your customers will all work differently.
- You also need to work out how all this affects their buying habits.
- List the reasons why someone may not be able to buy, and why they may be won’t buy.
- Gathering The Right Data
- Ensure that all goals/sales are being tracked.
- Look for backend data too, particularly if there are cancellation rates. If there are, speak to the customer support and sales team – ask what the customers are concerned about and how are you are selling the product etc.
Tools that can help with DISCOVERY:
UserTesting.com, Feedback Arms, Kiss Metrics, Qualaroo, Google Analytics, Crazy Egg.
Phase 2: Experiments:
- A Clear Methodology
- Have a clear hypothesis – what success looks like.
- Know the potential risks and communicate these to the client. Let them know there are risks involved – you need to prepare them to be patient and manage their expectations.
- Wireframes and Design
- Communication is key – show what you’re testing, explain what could happen and communicate what the results could be.
- Keep it on brand. Before you start, identify client brand and ensure they’re happy with what you’ll be doing for them.
- Ensure the new design is driven by your research and data – shouldn’t do the design based on anything else.
- Pre deliver what needs to happen and who needs to do it – everyone needs to be aware of what’s going on. Want things to be a smooth as possible. Want to do design and want it to be implemented as soon as possible.
Tools that can help with EXPERIMENTS:
Optimizely, Visual Website Optimizer, balsamiq, Google Analytics and Mockingbird.
Phase 3: Review:
- Data Analysis
- Was your hypothesis correct? What were you trying to prove in the first place? Shelve all other details, and focus on what you physically wanted to test.
- Check for anomolies – before you start drawing conclusions. It will screw up data if you don’t.
- Don’t keep hitting refresh.
- Roll out changes widely. Redirecting people isn’t good for the long term.
- Roll out changes widely – will need dev help for this. Redirecting people isn’t good for the long term.
Joe Doveton, GlobalMaxer – Bouncing Czechouts: Cultural Conversions and International Ecommerce:
You can view Joe’s deck, below:
How Can You Optimise Your Site For Different Countries?
- Do not use a ‘global template’, whereby you only need to design it once. The problem with this kind of design is that it may be easy to maintain, but it’s not the best approach for your customers.
1. “Greetings Earthlings”
- People combine languages around the world – whatever language they speak, many people always use a bit of English when searching for terms online as most people speak it – for example: Itenglian. When searching for cheap flights, ‘voli low cost’ has a higher search volume than the native ‘voli economici’. Therefore, localise, don’t translate – you customers have already adapted and absorbed an Anglophone web into their own culture.
- If going to do linguistics research, you need to use search engines that those countries use on a daily basis. For example, in South Korea, they use Naver instead of Google.
High Context and Low Context Societies:
Written text is only one part of communication – and research visual clues, colours and composition.
Need to take these into consideration when designing your website:
- High Context Societies – a lot of communication is implied and it’s up to the people to interpret it – ambiguity with the information you display.
- Low Context Societies – very much about the sole thing that’s being presented.
Hofstedes 5 Cultural Dimensions are important in line with this. Think about colours and their meanings in different countries – what colours are popular/ have a positive significance?
2. Eat Guinea Pigs, Don’t Eat Elephants
- Culture affects all facets of design.
- However, there are degrees of culture.
- Small is a tasty morsel – eat little and often. Culture is too big of a topic to bite off in one chunk. Pick a few battles you think you can win.
3. Bears in Space:
- You must do the technical research – It’s important to consider technology fashions – don’t assume your cultural customers have access to same technology as you, or use the same technologies on a day-to-day basis.
- EG: African mobile web – majority surf the Internet on mobile.
- Great video is no use in a country with poor broadband.
- Emerging markets may be behind or ahead of the technological curve.
- Need to also think about how your information is displayed – if building experience in Japan, you need to think about how to display everything depending on which language they use – two main languages to take into consideration (Tategaki – reads left to right, horizontal v. Yokogaki – reads vertical, right to left.)
- Therefore, be prepared to backwards engineer your web pages/sites/apps where necessary.
4. Chocolate Coins – Payments!
- Payment conditions differ widely between cultures
- EG: Arabic ecommerce – shut down your baskets when prayers are on.
- EG: India – cash on delivery preferred method of payment – big effect on returns policy.
- The relationship between the buyer and seller is also cultural
- There are also conventions to be respected in different countries about how you ask for business
- EG: China – how you ask for the business is important – they prefer more formal – “Request quote”, rather than “Get quote”.
Stephen Pavlovich – Conversion Optimisation for Mobile
- Industry we’re in is in a period of change – those companies embracing mobile are becoming more successful than their competitors, as they’re getting bigger conversion.
- Companies like Amazon, booking.com are gaining significant market share due to conversation optimisation.
- Recent survey from BBC News highlighted that PC sales are declining, but tablet sales are increasing. More and more people are using tablets and mobiles to buy things.
- Mobile commerce now accounts for 23% of online sales – differs depending on market but shows growing significance.
- Many companies are burying their heads in the sand and trying to ignore it. Most don’t have a mobile app and those that do aren’t split testing it or leaving it as basic. Investing in mobile will help you to grow long term.
- Plus.net is a good example – mobile and tablet saw 10x sales and 2x traffic.
- Bainst and Ernst another – 51% conversion increase. Got a case study mention from Google too – favoured in industry because of good conversion usage – advantage over competition.
12 Tips for Mobile Conversion:
1. Start Small and Scale
- You don’t need to spend a lot of money making a mobile site – there are ways to cheat. Pick one page on your website (landing page), split test it and then see the impact it has. If you see some lift, you have a good case to make a bigger mobile design.
2. Create Mobile and Tablet Dashboards
- Focus on sales funnel landing pages goals.
- Remember goals on mobile are different – when a user comes in on mobile, not all will convert into sale. May be trying to find out opening times etc – so be aware of different user goals and intentions.
3. Gather Mobile-Only Surveys To Gather Objections
- Overlay a simple survey on your mobile testing site, and ask users three simple questions:1. “What’s the purpose of your visit? We’re you able to achieve that?” Will show exactly why people are coming to your website. There will always be multiple reasons why and multiple reasons why they aren’t converting.
2. (At the confirmation page at checkout) “Was there anything that nearly stopped you buying from us today?” - Perfect question to find out what would be stopping users converting. Fix it, and you should see your conversion rate increase.
3. (If click “Switch to Website View”) “Can you tell us why you’re switching to the desktop site?”
4. Use Desktop Clickmaps to Priorities Content
- Use heat mapping to see where people are clicking.
- Have to prioritise content depending on where your users are clicking.
5. Dogfood Your Site on Mobile
- Use your own website and test it as if you are a customer – need a good understanding of their purchase intentions and behaviour.
- Use your own money to buy something rather than the company’s, so you understand same thought processes – you’ll become aware of all of the problems that your customers could encounter.
6. Identify Key Mobile Personas and Use Cases
- Identify three to five personas with usability tests, surveys and key word analysis.
7. Target Usability Tests on Potential Customers
- Usertesting.com – ideal way to find out how a regular person would use your website. You are not the right person to gain a fresh perspective, as you already understand the technical side of it.
8. Sketch Mobile Designs on a Business Card
- About the same size, so you can put content in the right order, and see if all of the important content fits onto it effectively.
9. …Then Create Wireframes in Axure or Adobe Edge Reflow
- Tools allow you to do responsive designs.
10. 3 Mobile Designs To Copy
a) AirBnB – call to actions are clear and tabs are present, so not bombarding with user with lots of text or images. Makes it easy to manage a lot of content on website.
b) The Home Depot – picks up user location straight away so it can show opening hours on landing page. Checkout has options to deliver or pick up from store with details of stock.
c) Target – big focus on images, rather than trying to cram text. Very simple. If you want to find out more info, you can but it won’t overload user. Checkout – actively tries to reduce amount of things you have to fill in to get the product. Higher conversion.
11. Prioritise and Group Mobile Tests
12. Iterate for Local Maxima, Optimise for Global Maxima
- Facebook option to deactivate an account good example:
- To reduce the amount of people trying to leave, Facebook focused on all of the reasons why you shouldn’t – “So and so friend will miss you”.
- They also lengthened the process to hinder you from leaving. Have to choose multiple reasons why you want to leave – Facebook saw significant in conversions as a result.
Philip Sheldrake, Euler Partners – The Value of Social
- Social is more than just selling:
- Marketing is about creating value for customers, not just about selling.
- It’s the process by which companies create value and relationships – all about giving the right product, to the right person, with the right need.
- Focus on outcomes:
- Rather than focusing on social metrics such as shares, likes, tweets, followers etc focus on outcomes.
EG: Although two companies may be doing the same things, the outcome and value will be very different from each company despite the output being the same à High metrics do not translate to good public relations.
- Just because things are easy to measure, or are already being measured, doesn’t mean they are necessarily worth measuring. Need to consider what measurement counts for.
- No universal metrics when it comes to social – all metrics need each other and no metric can achieve what’s needed alone.
- Never employ fewer or more metrics than will suffice. Handful of complimentary metrics work best.
- The effect on business results can and should be measured wherever possible:
- What does that mean for the business in terms of business outcomes?
- When presented with a set of data, ask “so what?” at least three 3 times in a row. If you can’t answer it, no-one will understand what all the effort was for.
- ROI vs Value:
There is a difference between ROI and Value in social. Value is far more important when carrying out measurement on social.
Ben Harris, Decibel Insight – The Analytics Renaissance
- Analytics VS Insight: Analytics is a confusing term – double sided coin:
- Analytics is about monitoring performance
- Insight is about improving importance
- It is therefore important to consider what information will be most useful in your Google Analytics account – 49% of people update their website on a weekly basis, but only 14% of people track this.
- Traffic, UX and Content make up the success funnel – the most important parts of a successful website.
How do you approach measuring UX? 6-Step Approach:
1. Learn about your visitors
No point trying to optimise your website if don’t know why your visitors are there.
Segment your traffic to the top 2% of visitors (those most valuable) – base your judgment predominately on this proportion.
Audience segmentation – De-skew data – you will have audiences with different needs, so you need look at the different ways to optimise for these different audiences – detect problems and then improve them.
2. Experience as your users experience
Make sure you do in-depth browser testing. View visitor playback to determine visitor navigation patterns, and carry out basis UX tests.
3. Find your most popular content to optimise user experience
Via heatmaps. You can then organise your content in such a way that your content is in an optimisation location. If can understand what content works, and understand how to do it better, can understand what converts well.
4. Ensure your visitors see the right content
Scroll map a good way to see how customers using page and what looking for – help you to move things and get better click through and conversion.
5. Understand how visitors navigate your website
6. Increase the effectiveness of mobile
Russell McAthy, Stream:20 – Excel, You Love To Hate It
As an Excel fan myself, rather than me trying to detail out every single handy formula that Russell dished out to us on the day, you can view all of his tips and tricks below:
Dara Fitzgerald, Fresh Egg – Attribution with Google Analytics – How to do it, and why you shouldn’t!
The full deck from Dara is available, here:
- Multi Channel Funnels (MCFs) in Google Analytics can help to give a better understanding of each channel’s role in conversion:
- Multi channel funnels have been in in GA for two and a half years now, and they help to improve performance online – 43% conversions are assisted here.
- However, due to multi devices, this figure is probably a lot higher.
- So to track conversions correctly, you need to have:
a) Conversion tracking (goals and ecommerce)
b) Ensure correct campaign tagging or the data is meaningless
c) Manual campaign parameters (URL builder)
MCFs need to be tailored to your business – can make tailored channel groupings, can alter to make sure match track.
However, key things to remember when doing so are:
- Don’t mix different activities together. Customise channel groupings in MCF as much as possible.
- Don’t treat all conversions as equal. Look at paths for different conversion types separately.
- Split true and self-assists for each channel and look at total channel contribution.
- Universal Analytics promises to greatly enhance possibilities for Customer Lifetime Value (CLV) measurement and optimisation.
- Run controlled tests to determine incremental value.
Nikki Rae, Future Insight Analytics – Getting Google Analytics Going From The Get Go
You can view Nikki’s full deck, here:
- In GA, what you see is not always what you get without customisation.
- If you haven’t got the traffic sources right in the first place, you’re not going to be making marketing decisions on the right things.
- “When is organic search not organic search?” Sometimes organic keywords can appear that you don’t rank for – this can be down to paid search traffic that isn’t tagged in Google Analytics from other search engines.
- “When is direct traffic not direct traffic?” Sometimes direct traffic isn’t direct traffic, it may be that Google Analytics doesn’t know where it came from. Campaigns may not be tracked correctly – such as email campaigns from Microsoft Outlook. Because Google can’t see where it’s come, it marks it as direct traffic.
- “When is a referrer not a referrer?” Your site may be a referrer to itself if your cross domain tracking isn’t correct, or search engines can appear as referrers when it doesn’t recognise it as a search engine, however a code can be added to change this.
- “When is a keyword not your keyword” Your tracking code is on someone else’s site, or you are viewing the Adwords Keyword report rather than the matched queries report.
Technical Check List:
- Make sure all of your pages are tagged – use SEO Screaming Frog to do so.
- Do an internal site search
- Ensure Webmaster Tools and Google Adwords are linked to the account to maximize output
- Ensure you only have one home page – this can be controlled using a search and replace filter
- Are you looking at the right traffic? Using filters, remove your staff and agency traffic via the IP address to check correctly.
Anna Lewis, Koozai – The Power of Segmentation in Web Analytics
Anna’s deck is available, here:
- We use segments to breakdown data to see the detail of it. Why? Because not everyone is equal, and it will allow you to see the full picture.
- Everyone that comes to your site does so with a different intent.
- Segmentation allows you to create personas to see what kind of user they are, and allows you to figure out how to market to different users depending on how you want them to behave on your site.
- 52% of digital marketers agree that the ability to personalise web content is fundamental to their online strategy.
- Can segment by demographics, conversion, persona, content viewed, interest level, changing activity, industry, traffic source.
- Google has got smarter with segmentation too – you can now segment traffic sources, ecommerce and sequences as well as users.
- It also allows you to import other people’s dashboards and segments.
- When segmenting, it is important to remember that is all based on a sample of data. So be careful about making big decision on small batches of data.
Ryan Gallagher, IOVOX – Real Time Business Insights
- The biggest challenge we face with big, important data is getting the most out of it and trying to make it as good as the human brain.
- Big data requires looking at a whole host of information and deciphering the most important aspects of it.
- Proxy phone calls allow you to get more from data and allows you to value leads more accurately.
- Can then use this to mapping trends, and look at external factors. For example on a sunny day, you’re more likely to get more visits than desktop visits, but on a rainy day, it’s more likely to be the opposite. Information like this can help you to decide when you should advertise, and what should advertise at particular times of the year.