Marketing Engineering: A systematic approach to harness data and knowledge to drive effective marketing decision making and implementation through a technology-enabled and model-supported decision process.
Market response models
Inputs: actions that the marketer can control + environment
Objectives: monitor/measure, etc.
Value = Benefits – Price
function (does the job)
economic (saves money)
subject and importance of need
temporal aspects of need (urgency, frequency and duration)
Measuring Customer Value
Objective Customer Value
Internal engineering assessment (internal estimate)
Indirect survey questions (pay for better quality, etc.)
Field value-in-use assessment (economic benefit): Not only marginal costs but initial investment, etc.
Perceptual Customer Value
Behavioral Customer Value
Statistical models / Data mining
Segmentation -> Targeting -> Positioning
how does the segmentation fit into the strategy
which variables can be used
how many segments?
reduce data (PCA)
develop measure of association
identify and remove outliers
form segments (cluster analysis): are they clear and robust?
profile segments & interpret
Attribute-based perceptual maps
identify other products & attributes
get data from questionnaires
Translating Preference to Choice
first choice rule: infrequent, expensive
Share of preference rule: often, cheap, etc.
For [target segment], the [offering] is [positioning claim], because [single most important support].
Sales force composite estimates: let the sales force forecast
Jury of executive opinion: different stakeholders
Delphi method: anonymous and iterative
Chain ratio method: split up in its factors
select attributes of product
segment customers based o part-worth function
design market simulation
select choice rule
adj market shares
Top 10 Lessons
Marking Engineering is Marketing
ME is a means to an end
frames the opportunity costs for alt. actions
whole greater than the sum of its parts
data & info do not automatically result in value
every model has its downside
ME requires lifelong learning
be a coach rather than a teacher
Insights for better Implementation
Start Simple; Keep it Simple
score inexpensive victories
develop a program, not just a project
I really liked
Principles of Marketing Engineering. The book gives a great overview over the topic and different approaches. It’s written like a textbook, so it can be a bit lengthy in parts but otherwise, a nice book.
Avoid “anticipointment”: bridging the gap from ad to site
Ads and web site work together – don’t just invest a ton in one medium
Marketeers fall easy into the ad trap because it’s easier than creating an usable, engaging web site
People expect that the click from the ad will be of even more value than the ad
Online Marketing Value Chain: Basically Customer Lifetime Cycle
Click ad, engage deeper in the landing page
Make their way through conversion opportunity
Become loyal customer
Most of these steps will be on the web site!
Creating a Successful Lead Nurturing Strategy, Part III: When Should I Call?
Call within 5 minutes of the initial contact
Call early at morning or late in the afternoon
Call on Wednesday or Thursday – I personally tried this against Monday and Friday and it was highly effective
Call them up to four times and send one email in the first 24h
Test these tactics
Creating a Successful Lead Nurturing Strategy, Part IV: Your Long-Term Strategy
The main is not to sell but to maintain a relevant conversation
Offer relevant and personalized content – recent study showed that most content simply sucks, so watch out
Email – automated, personalized and relevant; reports, tips, guides, best practices
Phone – Follow up; provide deeper information, answer questions
Direct mail – reinforce what you’ve talked about; again personalized and relevant
This process should be repeated maybe once a month
Optimization Is Greater Than the Sum of Its Parts
Testing & Targeting are greater than just once
however often they are siloed
Start with testing and segment the results
This helps you to find better content for targeting
Building a Business Case for Optimization
Biggest problems are processes and politics
They hadn’t ownership over the site
Testing generates positive ROI!
Optimizing landing pages increases off-site ROAS (Return On Ad-Spending)
Test to fail faster – some of your assumptions are probably wrong
Dig into analytics, segment and provide insights
The Collaboration of Testing Ideas
Include other people and departments in your testing
Often people in development, IT, creative, etc. have ideas – just ask them
Test different landing pages: home page, product page, internal search, etc.
Reinforce ad text/graphics on the landing page/multipage setup
Test incentives for submitting to your email database
Build a story with the ad and following pages
Test different viral/referral elements: coupons, vouchers, …
Test different forms
Test % Off vs. $ off
Test your CTA copy, size, color, style
Test scarcity on offers
Test different copy approaches: informative, funny, benefits oriented, etc. and analyze segment behavior
Test signs of trust: security message, shipping info, return policy, etc.
Test geographical targeting
Test simple content vs. rich media
Test content vs. no content
Test free shipping vs. % off vs. $ off vs. guarantee vs. …
Test promotion tresholds: 10% on $50 vs. 15% on $100
Test different internal search results – hand picked, automated, editor picks, big brands, cheapest first, best selling first, highest rating first, etc. and segment(!)
Understand your goal – what are you’re trying to improve?
Start with the bottom in your funnel – it’s easier to get more impact
Try to understand why alternatives work better
Try to improve one theme at a time, e.g. decrease registration drop off, copy style, etc.
Focus on big things: product shown, pricing, primary copy, images, offer, CTAs
Five Times to Test: 1 — When you need to optimize beyond the click
Data without analysis and communication is not very useful
Even then without taking action, it’s practically useless
Often lots of money is invested in driving traffic but less in converting the traffic
Example: large business $100MM PPC budget, less than $200k for optimizing landing page/website
Mark Typer, Wunderman: 15% Optimization, 85% Ad spending
Five Times to Test: 2 — To resolve internal disputes
Do you have a dispute? Just test the idea
Similar things can work different on different websites, e.g. CTA wording
Path Analysis: A Good Use of Time?
Customers like the path the want to, not the one you force
The tool don’t show which page in the path was most influential
Most tools don’t track the path correctly
Exception: Landing Page experience
Possible solution: Group relevant pages
Show most influential pages/li>
Make segmentation easier
Stop Obsessing About Conversion Rate
Overall conversion rate doesn’t allow for actionable insights
It only covers a small minority of all visitors
People are going to research on your site
People who want to learn about your company
People who need help with a product of yours
People that do something completely different
That is, you neglect a big part of your visitors
You’ll focus on short term gains
Task Completion Rate
Helps you to cover all customers
Successes outside from conversions
Ultimately understand your customers better
Getting Started With Web Analytics: Step One – Glean Macro Insights.
Understand the macro level first
1. How many visitors are coming to your site?
2. Where re they coming from?
3. What is the purpose of your website? What are your top three web strategies you currently working on?
4. What are you visitors actually doing?
Consultants, Analysts: Present Impactful Analysis, Insightful Reports
No data overload: Give value instead of data, provide recommendations
Tie your data to business outcomes
Use other data than just Clickstream
Don’t make it boring
Connect insights with actual data
Meet the “exceptions of scale”: If you are a big agency or written a book on WA, then people expect more from you
Do something unique
Paid Search Analytics: Measuring Value of “Upper Funnel” Keywords
Upper Funnel / Longtail Keywords can neglected because of the single session mindset
Understand each stage of the customer purchase life cycle
Map your keywords to each of those cycles
Measure success for each cycle differently
Category Keyword: Bounce Rate
Category / Brand: Time on Site
Brand: Visitor Loyalty
Conversion/Product: Conversion Rate / Leads
Aggregation of Marginal Gains: Recession Busting Analytics!
Often web analysts don’t focus on the
immediately achievable improvements Simple things:
Figure out where you are making money
Check errors in your email campaigns
Fix your top landing pages
Compare organic and paid keywords: Where are gaps between these two and why?
Ask your customer
Fix dumb stuff: e.g. check
25-point Website Usability Checklist
Analyze This: 5 Rules For Awesome Impromptu Web Analysis
Question: What would you change on this website?
Useful things to remember:
Don’t start with your opinion: you are a proxy for customers / visitors => Better: State hypothesis
Always offer alternatives
Offer data, even when you don’t have access to the site’s data.
State your assumptions about the site’s objectives
Focus on obvious and non-obvious things: micro and macro conversions, competitor’s site, customer satisfaction, ideas for testing, demographic trends, etc.
Web Analytics Segmentation: Do Or Die, There Is No Try!
Pick at least some segments in: Acquisition, Behavior and Outcomes
#1: Acquisition: Where does the company spent its most money on?
How to segment:
Context: How many visits?
(Optional): How many were new?
What was the cost of acquisition?
What value could we extract at a per visit level?
How many were able to accomplish their goal?
Was what the total value added to our organization?
What are the visitors doing?
What do people want to do on your site?
How deep are they browsing on your site?
How long did they take time till conversion?
How often did they visited your site?
=> Traffic source, conversion, average order value, actions, etc?
Loot at macro and micro conversions!
i.e. video clicks, adding products to wish list, applying for a trial, downloading white papers, etc. etc. etc.
5 + 4 Actionable Tips To Kick Web Data Analysis Up A Notch, Or Two
Go deeper — Don’t stop at the obvious border: compare off and online data, create CLV for ecommerce,
People against lonely metrics club Measure the complete site success
Don’t just report one month: at least three months, understand your business’s cycles, create annotations
Make insights in your data obvious: Better visualizations
Segment your data (previous summaries)
Don’t just look at the top 10 rows
Step away from one-session thinking (later summary)
Achieve multichannel analysis (previous summary)
Win Big With Web Analytics: Eliminate Data & Eschew Fake Proxies
Most reports are overloaded with data
Though, the readers want context and insights
Eliminate all useless metrics and data in your reports
Understand the desired outcomes
Rebel! Refuse Report Requests. Only Answer Business Questions, FTW.
Context is important – no business is the same
Answer business questions – what’s the driving request for the data?
Attributes of a business question:
They are usually open-ended and on a higher level
They need likely more information than just visits, bounce rate, etc.
They are seldom answered with tables
The Difference Between Web Reporting And Web Analysis
Data puking isn’t web analysis
Signs that you are doing web analysis:
Actions instead of data
Work with the business, measure economic value
Use the web analytics measurement model (previous summary)
You are doing a bit advanced statistics
If you work with targets
You provide context
You segment your data effectively
If you can provide an impact for a recommendation
If you use less than four metrics in a table
If you use multiple data sources
Again a great video, this time about Web Analytics by
Avinash Kaushik. I just love his no-BS style.
Ask the metric: So what? Three times, if it don’t give an action it’s useless
Data should drive action
Give people the information they need – don’t send them everything => no death by data
Home pages of websites, are no longer the home page you want
Where do people come from?
What are they looking for?
Context matters: previous months, years, etc.
Relative numbers more important than absolute numbers
Compare different metrics, e.g. conversion rate and page views
non e-commerce sites:
averages hide truth effectively
How often do they visit?
How recent did they visit?
Depths of visit
=> Understand the value: Loyalty
Survey people: What do they think about the content?
Bounce rate: Came and left
Segment by source, entry-page, landing pages, etc.
Rules for Revolutionaries
10/90 Rule: $10 Tools, $90 People: Understand Data & Business, Able to analyze => to extract value
Reporting is not analysis: Reporting -> provide data; Analysis -> prove insights
Data Quality can be low, but is still better than other data
Faith-based initiative: e.g. magazine ad without tracking
Make decisions, don’t argue about the quality of the data
Over time understand why quality is different -> confidence will get better
Conclusion Decision making is a journey, not a destination
=> Put some level of process in place, mostly for tasks, e.g. what happens to implement a test, etc.
if HiPPo (highest paid person’s opinion) makes the most decisions
=> make experiments
Learn to be wrong, quickly
=> You probably don’t know what your customers want