[Webinar] Optimise your website with Analytics & A/B Testing

Thursday, May 22nd

&

Introductions

A/B Testing?

Optimization tips

Live Demo

One-Click integration with AT Internet

Examples & Applications

Final remarks

Introductions

Introductions : Optimizely

Fabian Liebig

Marketing Manager

@Optimizely since June 2013

fabian@optimizely.com

@FabianLiebig

Dan Pete

2008 :

Dan — Campagne Obama

2008 – 2010 :

Carrotsticks, Spreadly,

et Y Combinator.

2010 : Création d’Optimizely

Avril 2013 : Levée de $28M pour s’exporter

Mai 2014 : Levée de $57M

$28M

$57M

Introductions : Optimizely

7000+ Happy

Customers

At internet

Introduction : AT Internet

18

+3 500

+ 350 000

+ de 150

years experience

customers from all sectors

sites measured

employees all over the world

At internet

Introduction : AT Internet

1. AT internet Tag is remotely implemented in MV test

2. MV test measures are restituted into the Digital

Analytics tool

1- AT tag

2 - Test data

Partnership: principles

Optimization Tips

1. Define & measure your most common visitor path

2. Identify webpages with highest amount of traffic

3. Investigate where you see the highest exit rate

4. Identify the pages that generate the highest revenue

5. Investigate which pages can provide most data on your

visitors

Optimization Tips

Case Study

Original Page Winning Variation

Case Study

+10% clicks

Case Study

Original Form Variation B

Case Study

Original Form Variation B

+116% filled out

forms

Results:

Live Demo

Platform Demo

The integration at a glance

The user sideIntegration in few clicks via the Optimizely visual editor

The technical side

Optimizely sends 3 parameters to the AT Internet servers

« s »: Account Number within AT Internet

« abmvc »: OptimizelyTestID[OptimizelyTestName]-0-VariationID[VariationName]

« type »: Type of Test, always « mvt » Multivariate Test

The AT Internet side« Usability » « Multivariate Testing » in left menu « Tests » get metrics on all tests launched on Optimizely,

classified per test, variants, and creatives.

The integration at a

glance

Source : help.optimizely.com

1. AT internet Tag is remotely implemented in MV test

Reduce tracking implementation time

2. MV test measures are restituted into the Digital Analytics tool

Consolidate all data for global analyses and data

interpretation

1- AT tag

2 - Test data

Partnership: benefits

MV testing results in your digital analytics

Optimizely will indicate the « Winner variation » according the

objective of the test

But is it also the winner according

your website objectives and KPIs?

Test results & site performance

Criteria for the test

= click ratio

Site KPI

= number of registration

winner

+10% clicks

Test results & site performance

Focus on a specific population: Segmentation

Site Segment

Sources

Conversions

Navigation

Etc.

Sources

conversions

Navigation

Etc.

segmentationSegmentation: How it works?

1. Get a deeper insight into test results directly within the AT Internet

interface.

• Sources, entry page

• Time spent, PV, navigation path

• Conversions funnel

• Purchase made

• Equipement , Etc…

2. Ensure there is no deterioration of your main or secondary

objectives

Validate a test upon your kpis

Examples & use case

Test on the message for an offer

2 for 1 -50%

Use case

Test on the message for an offer

2 for 1 -50%

Use case

50€ 30€

2 for 1 -50%

Test on the message for an offer

With Analyzer and our segmentation tool : check also performance of each

variation on indicators such as purchase amount, number of products

Use case

2 for 1 -50%

50€ 30€

Use case

Test on the message for an offer

With Analyzer and our segmentation tool : check also performance of each

variation on indicators such as purchase amount, number of products

Compare new visitors with returning visitors

are the results for new clients the same as for returningclients?

Understand distorted results

check if external event such as equipment (screens, browers..) can be the explanation

Go further than validate « a template »

check if the winner is also a winner for all type of topics or aisle products?

Other use case

Conclusion

2. ExperimentQueue and

Log of active experiments

3. RunExperiment

4. Resultsand Key

takeaways

1. Brainstorming & Prioritizing

Conclusion

blog.optimizely.com

help.optimizely.com

Fabian Liebig

fabian@optimizely.com

Case study

“PriceMinister integrates MV

testing and increases

turnover by 5%”

http://www.atinternet.com

http://blog.atinternet.com

Marie-Alix Achard

Marie-alix.achard@atinternet.com

&