2. If you only had $1.00 more to spend, where would you
put it?
• PR
• SEM
What drives this decision?
• SEO
• Billboards
• Radio ads
• Daily deals
• Street teams
• Flyering
• Direct mail
• Promotional discounts
• Adopt a highway
• Direct Sales
• …
3. If you only had $1.00 more to spend, where would you
put it?
Constant Channel Measurement
• PR
Everything else
• SEM Billboards
($25)
(-$3,000)
3%
• SEO Adopt a 4%
Highway (-$85)
• Billboards 5%
• Radio ads
• Daily deals Guerilla ($87)
8%
• Street teams
• Flyering SEM (-$25)
10%
• Direct mail Radio ads
($100)
• Promotional discounts Word of mouth
58%
($95)
• Adopt a highway 12%
• Direct Sales
• …
When you’re young, do what you’re good at
4. How we learned where to spend on acquisition
Build a strong brand identity
Who will use us?
Ideation
What do they need us for?
Where will they realize they need us?
Test new channels Reinforce working campaigns
Test
Test new campaigns Scale effective channels
Positioning Top of funnel
Target market CPA
Refine Measure
Use cases Churn
Marketing channels Lifetime Value
5. How we learned where to spend on acquisition
Build a strong brand identity
Who will use us?
Ideation
What do they need us for?
Where will they realize they need us?
Test new channels Reinforce working campaigns
Test We measure
Test new campaigns Scale effective channels
impact every day
Positioning Top of funnel
Target market CPA
Refine Measure
Use cases Churn
Marketing channels Lifetime Value
6. How do we measure? Essential tools for startup
marketers
• KISS Metrics
• Google analytics
• SurveyMonkey
• GUI based email provider (madmimi, mailchimp)
• SQL
• Shell based scripting language (python, ruby, etc.)
• Graphic design: Photoshop, Illustrator
• Rapleaf
• Daily marketing dashboards (Excel, RJ Metrics, etc.)
• Interview real customers in person
7. Go beyond: Our internal tracking tool gives us more
visibility
• Tracking by user Today First Task Posted
– Parse HTTP requests
– Referrer, utm
Email Newsletter
parameters, domain, search
term
– Associate with user ID
– 30-day retroactive cookie
User ID 228297
Date 2/28/12 3:52
utm_source google User signs up
utm_medium cpc
data http://www.google.com/aclk?sa=l&ai= …
domain www.google.com
utm_content FurnitureDelivery_BP
utm_term furniture delivery service
utm_campaign S_Tasks
Set tracking cookie
This user may have joined from 30 days ago
adwords
Identify macro-trends in channels, messaging, demographics, zip codes
8. Automatic channel surveys reveal the qualitative
• 60% of all new TaskPosters receive a quick survey
• Helps measure word-of-mouth, offline and other hard-to-
quantify sources
10. Don’t forget about retention – it’s cheaper
• Email newsletter weekly A/B testing by geo
• Rules and behavior based emails
Test
• Loyalty program user research
• Advocacy and research groups Measure
(e.g. Yelp elite, ModSquad)
11. Weekly Testing: Newsletter example
• Current “Rabbit Report” converts at 20%
• Let’s test 30-day conversions per user
– TaskRabbit vs. Task of the week
View rate: 40% View rate: 40%
Cost per email: $0.01 CTR: 60% Cost per email: $0.01
CTR: 80%
Cost per task: $0.09 Conversion: 10% Cost per task: $0.42
Conversion: 35%
12. Be Holistic
• Test everything: not just channels
• It’s brand, location, messaging, value proposition, AND
channel testing
Geo
5%
Disp.
Conv.
10%
Disp.
Conv.
Positioning
Channel
Brand
Hinweis der Redaktion
Stand for somethingLanding page?
Stand for somethingLanding page?
You have elements of an entire company contained within a single person or small group of people: Part product manager, part creative director, part analyst, part engineer, and part user researcher
Hyper local matters
This isn’t just a 1-day click test, it’s a 30-day attribution by user ID that makes our measurement techniques unique