There is so much data, and so many metrics out there it can feel overwhelming when trying to decide how to measure the growth of your business.
In this SlideShare, we walk through the frameworks and concepts that will help you measure and grow your startup to success.
4. Peter Drucker, regarded as the founding father
of modern business management, commented
that itâs extremely hard to implement a new
business strategy within a company if the
existing culture doesnât agree with it.
Company culture isnât about ping-pong and
free burritos, itâs how your team tackles
problems. Do they make decisions on gut
reaction or take a scientiïŹc approach?
As a startup, youâre at stage where you can
shape the culture. If you want to create a
culture that makes data-driven decisions then
now is the time to do so!
@HUGHHOPKINS
5. When was the last time
you changed your mind
based on data?
@HUGHHOPKINS
6. The Lean Startup methodology by Eric
Ries is one of the inïŹuential movements
in how startups tackle problems. It relies
on creating feedback loops so that quick
and meaningful iterations can be made.
Measuring, Data and Learning are a
critical part of the loop.
Meaningful metrics are metrics that
enable you to make decisions.
@HUGHHOPKINS
11. @HUGHHOPKINS
Vanity
Donât help you make
decisions.
E.g. total signups. The
number will always grow.
You had 5,000 signups last
month and now you have
6,000 this month. It looks
good but isnât actionable.
Actionable
SpeciïŹcally ties your
actions to an observed
result.
E.g. the ratio of signups
that activated this week.
13. @HUGHHOPKINS
Happen after the user
activity.
Often easier to measure.
E.g. Revenue is a lagging
indicator of future growth.
Lagging Leading
Activities that will lead to
future outcomes.
Often harder to measure.
Hard to tie to any outcome.
E.g. Net Promoter Score
(more on NPS later) â high
NPS is often a leading
indicator of revenue growth.
14. @HUGHHOPKINS
Granularity vs volume
How long is your sales cycle? If it takes 3 months to close a sale and there
are only 10 opportunities in the pipeline then daily metrics on sales are going
to erratic and misleading.
Itâs important that the timeframe and volume of data in which metrics are
reported make sense.
If the metric is erratically changing day to day from 0% growth to 100%
growth you probably need to look at the metric on a longer timeframe. If itâs
too ïŹat then perhaps you need to look at the metric on shorter timeframe.
16. @HUGHHOPKINS
Metrics you need to track
CAC - Customer Acquisition Cost
CLTV / LTV - Customer Lifetime Value
CAC > CLTV = đ
CAC < CLTV = đ
The problem is that as a startup you donât have a clear understanding of
CLTV (you just donât have the data) but you also donât have a clear idea of
CAC too. The important thing is to track these costs. If CLTV is not
signiïŹcantly higher that CAC (i.e. 4x) you might be in a bad spot.
17. @HUGHHOPKINS
AOV - Average Order Value
CR - Conversion Rate
RPC - Repeat Customer Rate
Shopping Cart Abandonment Rate
Remember to segment these metrics by product, channel and
cohort. The act of segmenting each metric will help you spot
strong and weak areas of your funnel.
For example, breaking down CR by channel might reveal that
Pinterest performs 10x better than Twitter.
20. @HUGHHOPKINS
By Dave McClure of 500 Startups focuses on the 5 steps of the
customer lifecycle:
âą Acquisition
âą Activation
âą Retention
âą Referral
âą Revenue
Or âAARRRâ hence why itâs called Pirate Metrics.
32. @HUGHHOPKINS
By Alistair Croll and Ben Yoskovitz they advise focusing on the
One Metric That Matters (OMTM) right now to your business. You
should move methodically through the 5 stages of growth:
1.Empathy - IdentiïŹed a problem that user will pay for.
2.Stickiness - The product has the right features/value retaining
users.
3.Virality - Uses and features fuel growth.
4.Revenue - Finding a sustainable and scalable business.
5.Scale - Scale mass market or successful exit.
45. @HUGHHOPKINS
Cohort analysis allows you to compare how different groups
perform. The most common is to compare different groups of
users based on when they signed up.
This is particular valuable to startups. Again, everything is in ïŹux,
the onboarding ïŹow that user experience in January will be
different to the one users experienced in July. You need to use
cohort analysis to understand which one performed better.
52. Naming conventions
@HUGHHOPKINS
Was the metric name âaccount createdâ, âCreated-Accountâ or event
âaccountCreatedâ?
An inconsistent naming structure will lead to confusion in the team,
sometimes referencing the wrong metrics.
There are many naming conventions but the important thing is to pick one
and ensure the whole team follows it. i.e. âobject - actionâ all in lower case.
53. Transparency
@HUGHHOPKINS
The opposite is a culture of secrecy and how can you ïŹx problems that
arenât visible. Use that old laptop to power a large display showing your key
metrics. Youâll be amazed by how much team members will pick up from the
metrics, spot anomalies and be quicker to act in a change of numbers.
Being transparent and clear also prevents confusion. Ensure you have a
clear deïŹnition what each metric is. A common are of confusion are Daily/
Weekly/Monthly Active Metrics (WAU), for example is an active user deïŹned
as just logging in or using feature x, sending more than 10 messages.
54. Track from day one
@HUGHHOPKINS
Too often analytics is an afterthought when building a product.
Tracking and analysing the data is an essential part of the feedback loop.
When building products think how user behaviour can be tracked, what
metrics can we track that will enable us to make decisions.
When planning to build a product or feature plan how data will be tracked and
how that will help form decisions.
55. Automate
@HUGHHOPKINS
No one likes manual data collection. A manually updated spreadsheet will
work for the ïŹrst few weeks but eventually it will go out of sync as colleagues
will forget to update it.
Automate the data collection. If youâre focusing on a few key metrics make
sure the collection is automated.
56. Bugs đ
@HUGHHOPKINS
If a metric looks off double check itâs not a bug!
In a fast moving startup changes are being made daily to the codebase, how
customers are being acquired and onboarded are also changing. As a result
bugs naturally creep in and metrics will be skewed.
Double check metrics from time to time, sign up to your own service,
purchase a plan and check the numbers add up.
57. User testing
@HUGHHOPKINS
Startups, being young, have limited data sets in which to make decisions.
User testing is one of the most effective forms of testing you can do for your
startup.
It will help you spot bottlenecks and areas of confusion for the user in
minutes.
A good tip to bake user testing into your workïŹow is to commit to doing a test
on the last Friday of each month. Add it to your calendar and stick to it!