How to build a data-driven marketing strategy

Building a marketing strategy is like building a nation with proper frameworks that can be scalable and also sustainable. So cracking that is not going to be easy. Finding new ideas and experimenting assumptions is part of a digital marketer job. The inspiration of this write up comes from a recent I event attended from Hubspot “Five ways to build a $100 Million business”. Most of the information they shared is basically on the foundation of marketing channel you need to choose based on your company type but they reminded about a framework to systematically crafting your marketing strategy with CLEAR TARGETS & EXPECTATIONS.

First, Why do we need to have different marketing strategy/framework?

Because each business type is different based on the size of their target segment,  So we need to understand things like:

  • Which marketing channel to use (Offline – TV Ads, Newspaper, Corporate Events sponsoring, etc ; Online – online webinars, Social Ads, Search Ads, SEO, Content marketing, etc.)
  • The time from the first exposure of a brand to successful deal/ successful transaction VARIES for every business . (For example, an on-demand startup like Uber need few hours to convert someone it’s customer but whereas a big deal like property buying or enterprise software takes more than months to make a user into consideration of their brand)
  • The cost per lead signup or a conversion rate from leads to successful deal/transaction is going to be different as well.
  • The journey of a user can be multiple interactions OR single interaction (For example: To buy a coconut, you go to a roadside stall and buy it (single interaction). To buy a car, you check online reviews, friends recommendation, multiple advertisements which pushed to you, few sales calls, finally few on-shop meeting, etc)
  • Average order value and Lifetime value is going to be different, thus going to be different margin > That impacts how much you can spend per signup.

As you see there is NO one size fit for everything… So how?

In this below example which inspired from Chris article, let me put in a table view.

What type of company you are? Elephant or Mice? How much customers you need?

If Type of company	Example Company	It’s Lifetime revenue(LTV*) per year per customer	Total Number of Paying Users Needed	Forecasted Revenue = Multiply LTV * Total paying users	At 30% (optimistic) Conversion rate from signup to transaction, how many leads we need Elephants (Service to Enterprise)	Salesforce(software selling) 	$100K	1000	100 Million	3,300 Leads we need Deers (Medium Startup)	Hubspot(software selling)	$10K	10K	100 Million	33,000 Rabbit (any consumer)	An Airline	$1K	100K	100 Million	330,000 Mice	Uber	$100	1 Million	100 Million	3,300,000 Flies	Facebook	$10	10 Million	100 Million	Just run ads for all users

(*simple definition of Lifetime revenue = total amount people give to us from customer wallet in a one-year time span)

Using this above table as an example, let’s systematically define how much we money we need for marketing or say how much we should spend to get a user and how the conversion rate of each journey should be (This target is important to make profit for every acquisition we make)

UBER: An Example Framework

  • Uber whose average transaction cost per user (usually my trip cost around $10) and If I take 3 times a week * 12 months > Total Lifetime value from JK to Uber = (10*3*12) = USD360.

How much I need to spend per paying customer? (CAC)

	Uber Average Transaction Value	10 Number of Rides	3 per week Transaction value on year1 v(Rides*Week*Months=10*3*12)	360 Assume every user stays averagely 	1 Year Lifetime value	360 Uber Margin	20% (assumption) Gross profit	USD72 Operational cost (Non-Direct Ads spend)	50% of gross profit Net profit (minus of 31).                        	USD31 So how much we can spend per user for direct Ads spend?	USD31

So if we spent more than USD31 on any ads, the company going to lose money. So if they spend only $15 per paying user which means $16 profit per year 🙂 Anyway the point here is when you run your own company or taking care of marketing, this overall marketing cost NEED TO BE tightly monitored and it shouldn’t go above this USD31 at any cost.

So what should be our leads cost? (signup user or CPA)

Cost per Paying User Target Cost per click on Singapore on Facebook  Conversion rate from traffic to lead signup on landing page For every $100, we spent Per 100 Traffic, so  Cost per signup Per 30 leads, with 40% conversion rate to successful transaction So how much we spent? Again How much paying customer we got? So cost per paying customer? which is less than $31 target, is it profit? How much?($31-$8)
Calculating how much the signup leads cost should be


Anyway this is a too good number. Meaning conversion rate from traffic to lead is not going to be easy! 🙂 and Conversion rate from signup to successful deal is not going to be EASY! 😀 This is where the efficiency comes into mind.

  • Improving the landing page conversion rate (on user boarding things)
  • Improving the email marketing campaigns that can nurture from lead to successful transaction, etc.

Ok now we know how much we can spend to acquire a new customer and new lead. Let’s

Which marketing channel is good for me?

  • Do you think Uber should sponsor a luxury resort to sponsor for a event where 1000 people going to come? (Sponsor cost = 50,000 ; Assume 1000 people also going to be a paying customer , so 1000*$31(LTV net profit) = 31000 . Is it a profit or loss? ) Should uber continue this or not?
Finding how much budget we need per channel and forecasting the impact

Ok now we need more channels to scale… Assume it is recruiting influencers to write about us.

Apart from all this numbers, the basic requirement is (a) the product fits the market and people need it (b) the market size is more than 1million if you want 1Million users

So if the basic requirement is clear on the product and market segment, then specifically when you run marketing on online, you should be tracking this numbers this way and have PROPER TARGETS for each layer. One affects the other in the chain. yes, it’s like a chained machine which depends on each, if we make a positive improvement on any joints, we can see improvements on the whole chain…

In conclusion:

  1. So being aware of each target, expectation of how much signups we need, and how much conversion rate we need to maintain, we will be able to planand
    1. To decrease cost per signup
    2. To increase conversion rate
    3. To increase lifetime value of customer
    4. How much budget we need for marketing, etc.
  2. Different marketing channels play different for every companies, so we need to know What marketing channel we need
    1. How much time each marketing channel can take a user to convert from ad exposure to successful transaction
    2. How much effort we need to put for each channel

Hope you found this information useful to plan you target framework and marketing Channel strategy to be successful in what you do! IF there are corrections in numbers, kindly let me know… Am good at making mistakes with numbers 😉

Reference: Hubspot startup Event & Further research: http://christophjanz.blogspot.com/2014/10/five-ways-to-build-100-million-business.html

2 replies to How to build a data-driven marketing strategy

  1. Pingback:

  2. Pingback:

Leave your comments