The Online Acquisition Framework

In my last article I’ve written about how to find out if a business is scalable by examining the economics of the sales unit, i.e. cost & revenue. From the examples demonstrated, we’ve learned that the revenue performance are entirely dependent on the acquisition type of the business, and to recap:

  • For customer-centric businesses that acquire customers, the revenue performance will be determined by 4 variables, i.e. average customer churn rate, average transaction frequency, order value, & margin %.
  • For transaction-centric businesses that acquire transactions, the revenue performance will be determined by only 2 variables, i.e. average order value & margin %.

Nonetheless, regardless whether the business is customer-centric or transaction-centric, the cost performance are both determined by the same variables, i.e. total cost to win acquisition & total acquisition. Which brings me to my next point that I’m dedicating this article to deliberate on acquisition, specifically in a context of us living in the information age where consumers are constantly experiencing information overload – how would we know what drives the acquisition along the customer journey? And how do we make sure we are investing the acquisition cost in the most objective manner?

Identify your online acquisition funnel stages

 
First and foremost, any businesses that operate in a free market are bound to compete with the market supplies before it can successfully acquire any sales. The fact that we, human as an emotional being that is constantly swayed by the information we consume prior to making any decision, had established the principle that every sale (acquisition) intrinsically resulted in a funnelwhere the drop-offs are inevitable due to the competition in this free market.

The reality about the acquisition funnel is that the drop-off does not just happen in one occurrence, but rather in a cascading manner throughout the timeline of a funnel journey. Hence, to strategically minimize dropping-offs throughout the funnel we should identify which are the stages that are prone to encounter drop-offs along the funnel, and minimize the drop-off rate for each of the stages. In other words, instead of only assigning one goal for the entire funnel, it is rather strategic to break it down into individual goals for each stage where the drop-offs are prone to happen, to have better insights and conclusive plans to improve your entire acquisition funnel performance. 

Let’s take online store acquisition funnel as an example, throughout the online shopping journey if there were 100 users visited your product page and ended up with 20 successful transaction, it would be too generalized to conclude that the product is the reason that resulted in 20% conversion rate; or it would be too generalized to conclude that the product page is misleading that causing the drop-off; or its also generalized to say both the product & product page are not the culprit, instead the initial ads content is the reason that resulted in 20% conversion rate. One way or another, without the funnel stages being identified we could not learn actionable insights as to how to improve acquisition effectively.
 

Adopt a ‘single source of truth’ framework to evaluate your online acquisition performance

 
Being an emotional creature, every bit of information we consumed will incrementally shape our decision-making, including our purchase decision. Hence, the media has been the long-standing propaganda tool not only to earn public opinions but also to earn market share. Not to mention the impact of online media as we currently live in the information age that are constantly served by infinite information distributed digitally during our day-to-day internet consumption. Say, you have an online business, and you are also investing in online media to earn market share, how do you find out which online touchpoints on the World Wide Web that you had invested shaped your customer’s buying decision, from the very first encounter with your store (website), till the moment they finally completed a payment
 
  1. First, you will need a neutral framework to record all visits to your store (website). It means you will need one SINGLE SOURCE OF TRUTH where every visit coming from any sources on the World Wide Web is collected in ONE single platform so that all data are evaluated EQUALLY.
  2. Secondly, to be able to evaluate touchpoints that ONLY fall within the customer purchase journey (that is, from their very first encounter with your store, to the moment they completed payment), you will have to find out what is the typical time frame of your customer purchase journey.
  3. Thirdly, to be able to evaluate your customer digital touchpoints by individual level instead of aggregated level, this single source of truth will need to assign a unique identifier for every user that visited the store.
Despite the fact that this is not a sponsored article, it is fair to point out that Google Analytics has been the go-to web analytics platform that ALMOST fulfills the requirements for being the single source of truth for your online acquisition framework. What’s more, it is a free-to-use tool as long as it doesn’t exceed 10 million hits per month. 
 

(I) How ‘Users’ are defined by Google Analytics

 
Having said that, you may have to bear in mind that Google Analytics are not able to automatically assign unique identifiers to individual users per se. This is because by default, Google Analytics are only assigning unique identifiers for devices or browsers. In other words, it could not tell if it is the same user that visited your store if he/she has been using different devices (laptop/mobile/tablet) at different times. To overcome this, you could work with your developer team to implement User-ID in Google Analytics tracking code on your store – where every time your visitor sign-in your store, the authentication system will auto generate a user ID and send it to Google Analytics to be associated as a unique user. As long as this advanced implementation is not in place, your default ‘user’ identifiers on Google Analytics are actually associated with the users’ device or browser.
 

(II) How are online touchpoints being reported by Google Analytics

 
By default, Google Analytics record the source & medium of every touchpoint (visit) to your website from the World Wide Web, whereby:
  • source is the website where the visit came from
  • medium is the medium type of the said website source
However, when it comes to reporting these source & medium, the data that you will see on Google Analytics default reporting dashboard are in fact the last touchpoint that users have ever clicked on, and Google Analytics by default do not report ‘Direct’ as the last touchpoint source. I believe the reason that Google Analytics by default ignoring direct touchpoint as the source, is so that marketers are able to learn external touchpoints performance to the greatest extend. If it happens to have visit coming directly to your website without any external sources at the point of visiting, Google Analytics will trace back its history (within the timeframe you set as your customer purchase journey, e.g. last 30 days) to check if this user (or rather device) had visited your website via any other touchpoints. If it’s found that this particular device had visited your website in the last 30 days before, it will then take the last touchpoint to substitute this ‘Direct’ source instead.
 
Speaking of customer purchase journey, as mentioned earlier it is recommended to set a realistic timeframe of which you think your customers would take to complete the full journey, i.e. from first encounter with your website, to the moment they completed payment. Take note that by default Google Analytics set this time frame to 6 months:
It means that Google Analytics is assuming your customer purchase journey is 6 months, whereby after the first time your prospect visits the site he/she should come back revisit again and make a payment within a 6 months time frame. To be able to evaluate your online touchpoints performance objectively you may want to adjust this time frame depending on your products type. For example the purchase journey for a face mask may only last for a week, or less. If the prospect’s first touchpoint was on 1 Feb but did not make a purchase, then visit your website directly after 5 months to buy a face mask, it may not be justifiable to give credit to the first touchpoint that happened 5 months ago.
 

Now that you know how online touchpoints are reported by Google Analytics, let’s elaborate further with a scenario where you are running an online store that sells face masks, and you presume your customer journey timeframe to be 7 days. One fine day there are 9 online users who have not visited your website before, came across your online store website on Google search organic result, clicked your website link and landed on your website. On the same day, there’s another online user who visits your website directly via www.yourwebsite.com; this particular user is not new to your website, as he/she had visited your website before via the Google search organic result touchpoint 6 days earlier. On Google Analytics it will see that there are 10 touchpoint sessions reported coming from 10 online users, where the source is ‘google’ and the medium is ‘organic’:

10 touchpoint sessions by different users (devices) reported coming Google search organic result

On the surface it may seem as though all 10 touchpoint sessions came from Google search organic results, although there is one session that literally came from direct visit. Once you add a dimension ‘direct session’ next to the source/medium you could then find out there is one session in fact came from direct visit, but due to Google Analytics default ‘non-direct attribution’ rule, it is not reported as the source on the surface.

 

(III) How to know if the online touchpoints have been fueling your acquisition funnel in stages

 
Presuming that you are investing in all sorts of online media for your online acquisition, e.g. social media, search engine, instant messenger, video, email, blog, e-news, online classifieds, display networks, forums, webinar, podcast etc, you will want to be able to identify which specific touchpoints had fuelled your customer purchase journey to justify which touchpoints deserve more investment along the way. Especially knowing that each ads buying tool (e.g. Google & Facebook) would attribute conversions to their ads respectively even though users may have interacted with ads in other channel in the same period of time. Hence, this is when the single source of truth come into play – first of all you will have make sure you have the following in place:
 
  1. Your customers are able to perform the entire purchase journey within just ONE website domain of yours. In the event where you need to divert users to another website domain to complete the purchase journey, you will have to make sure both websites share the same of your Google Analytics tracking code, and have implemented cross domain measurement.
  2. As established earlier that acquisition funnel dropping-out happens in a cascading manner instead of in one occurrence, you will want to set up Google Analytics goals according to your acquisition stages. This will not only help you to evaluate which touchpoints have been fueling each individual acquisition stages respectively, it also helps you to be specific on which goal you want to drive in your ads buying tool (Google or Facebook).
  3. Once you have the Google Analytics goals set up, you will want to make sure all touchpoints that you are investing are clickable via your website URL, so that it will direct user to your website where you have the Google Analytics implemented, and subsequently report which touchpoints had completed the goals that you had set up in the first place.
  4. The website URLs that you embed for each touchpoints should be unique from one another, whereby the URLs are being tagged with utm codes that represent the touchpoint’s attributes. The utm codes consist of 5 parameters where ‘source’ & ‘medium’ are mandatory to embed, while ‘campaign’, ‘content’ & ‘term’ are optional to embed if marketers want to know more specifics of the touchpoints.

For example, in this coming CNY season you are selling CNY hampers on your website, and you’re planning to attract customers from your existing email database, associate’s newsletter, affiliate website banner, Instagram profile, Facebook post, Facebook catalogue ads, Google shopping ads, Google search ads & Youtube video, Whatsapp message & Telegram message. This is how you may build your touchpoints URL with utm codes.

Touchpoint destinationTouchpointSourceMediumCampaignTouchpoint URL with UTM codes
www.yourwebsite.comYour customers newslettercrmemailcny-hamperwww.yourwebsite.com?utm_source=crm&utm_medium=email&utm_campaign=cny-hamper
www.yourwebsite.comAssociate’s customers newsletterassociate-website-nameemailcny-hamperwww.yourwebsite.com?utm_source=associate-website-name&utm_medium=email&utm_campaign=cny-hamper
www.yourwebsite.comAffiliate’s website banneraffiliate-website-namebannercny-hamperwww.yourwebsite.com?utm_source=affiliate-website-name&utm_medium=banner&utm_campaign=cny-hamper
www.yourwebsite.comInstagram profileinstagramprofilecny-hamperwww.yourwebsite.com?utm_source=instagram&utm_medium=profile&utm_campaign=cny-hamper
www.yourwebsite.comFacebook postfacebookpostcny-hamperwww.yourwebsite.com?utm_source=facebook&utm_medium=post&utm_campaign=cny-hamper
www.yourwebsite.comFacebook catalogue adsfacebookcatalogue-adcny-hamperwww.yourwebsite.com?utm_source=facebook&utm_medium=catalogue-ad&utm_campaign=cny-hamper
www.yourwebsite.comGoogle Shopping adsgoogleshopping-adcny-hamperwww.yourwebsite.com?utm_source=google&utm_medium=shopping-ad&utm_campaign=cny-hamper
www.yourwebsite.comGoogle Search Adsgooglesearch-adcny-hamperwww.yourwebsite.com?utm_source=google&utm_medium=search-ad&utm_campaign=cny-hamper
www.yourwebsite.comYoutube videoyoutubevideocny-hamperwww.yourwebsite.com?utm_source=youtube&utm_medium=video&utm_campaign=cny-hamper
www.yourwebsite.comWhatsapp messagewhatsappmessagecny-hamperwww.yourwebsite.com?utm_source=whatsapp&utm_medium=message&utm_campaign=cny-hamper
www.yourwebsite.comTelegram messagetelegrammessagecny-hamperwww.yourwebsite.com?utm_source=telegram&utm_medium=message&utm_campaign=cny-hamper

There are plenty of free tools that could help you generate these UTM links automatically by just inserting each parameter value, for example this Google Analytics UTM builder chrome extension. Note that these parameter names are fixed, we can’t change the parameters to another name because Google Analytics will not register any other parameter names other than these preset ones. But it’s totally up to you on how you want to make use of these parameters because eventually you are the one that will make sense of the attributes you set in the first place. For example, you may not want to monitor the effectiveness of ‘content’, because the ‘campaign’ parameter pretty much sums up the content of the touchpoints. Instead you may want to monitor audience type effectiveness, especially when you think investing in the audience who are ‘similar to whoever that had completed payment on your website before’ could be more cost efficient than the audience without specific targeting criteria. In that case you could fill ‘content’ utm parameter with ‘audience’ attribute instead, and moving forward use that as your reference for audience type.

Once online users come through your website via these touchpoints URLs with utm codes, Google Analytics will report all the touchpoints performance according to goals. As you can see from the example screenshot below, each touchpoints’ source, medium and campaign attributes are being reported in rows, and you can even choose individual goals from the right top drop-down to see how each touchpoints perform according to each funnel stages (goals).

If you ever notice there’s ‘Direct’ touchpoint source being reported here, it would be due to either of these three reasons:

    1. These online users have never clicked through your website via any of your touchpoints at all, he/she just got into your website by typing in your website address on the browser.
    2. These online users came across your touchpoints before, but it was long before the expected purchase journey that you had set in the first place. For example if you have set the time frame to 30 days, the user came to your website via your Facebook ads on day 1, but never came back again until on day 31 he/she came back for the second time via your website address directly.
    3. Because there are certain touchpoints not being tagged with utm codes properly, or not being tagged at all. I’ve attached a video below as one of the examples when the email touchpoint is not being tagged with utm codes, end up reported as ‘Direct’ source on Google Analytics.

 

Use this point of reference to control your acquisition cost

 
Now that you have the single source of truth (SSOT) for your touchpoints investment, it will be your point of reference to control your acquisition cost, given that most of the touchpoints that you are investing will incur expenses in either per-thousand impression, per click, per view or per post basis. Regardless of the expenses on which basis, you would be able to objectively evaluate each cost per acquisition (CPA) via this SSOT framework, because each acquisition (referred as goal on Google Analytics) is being reported according to touchpoints. 
 
For example, you know beforehand that once an acquisition happens, you will be generating $100 gross profit. So when you have this SSOT framework in place, by pulling the total ad spend from each touchpoints channel and referring total acquisitions accredited here you will be able to tell which touchpoints able to keep your CPA within, say, just 30% of your gross profit, which is $30 even though each touchpoint charges differently.
 

One way to optimize your acquisition cost efficiently is by automating the ROAS (return on advertising spend) rule if it’s available on the touchpoint channels you are investing. Currently, the digital duopoly Google & Facebook are equipped with this ad buying automation where you can command Google or Facebook ads to generate the returns ratio that you want to achieve against the ads spend. In order for Google ads & Facebook ads to automate the returns ratio that you would expect, you will first need to ensure both Google ads & Facebook ads have the information about your transaction value. That means, every time when the ads generate sales, your website is able to send the transaction amount back to Google & Facebook ad buying tool to reverse-engineer the return & ads spend ratio. You can send transaction value back to Google Ads by setting up transaction-specific conversion tracking on your website, whereby you will need to work with your web developer to push transaction value dynamically. Whereas for Facebook ads, you can send back transaction value by setting up Facebook pixel with ‘purchase’ event tracking on your website, also you will have to work with your web developer to pass the transaction value to the pixel dynamically

This is where you can set target ROAS on Google ads

(Only available when your Google ads campaign generated at least 20 conversions in the past 45 days, except for Search campaigns, which need at least 15 conversions in the past 30 days.)

And this is where you can set minimum ROAS on Facebook ads

(Only available when your Facebook ads generated at least 30 click-through purchases over the last 7 days)

For example, you want to generate 300% returns from your ad spend. In Google ads you could key in 300% in the ‘target ROAS’ field; while in Facebook ads you could key in $3 in the ‘minimum ROAS’ field, that’s telling Facebook for every $1 ad spend you want to expect a minimum of $3 sales. However, take note that this ROAS automation does not factor in the CPA value that you want to achieve. This rule is only telling the algorithm to focus on generating the returns ratio while your target CPA will be ignored. 

Also note that your acquisition cost performance should not be limited to just the pricing factor of the touchpoints channel. You will want to scrutinize the conversion rate for each touchpoint carefully too, because you could have the same budget allocated across all channels, with the same pricing, but the conversion rate will determine if you are achieving the cost per acquisition effectively. Let’s illustrate with an example below:
 Channel AChannel BChannel C
Budget allocated$1,000.00$1,000.00$1,000.00
Average Cost Per Click$0.50$0.50$0.50
Total clicks200020002000
Total conversions302010
Conversion rate1.50%1.00%0.50%
Cost Per Acquisition$33.33$50.00$100.00

As you can see, even though the budget and CPC pricing are the same across all A, B & C channels, but with different conversion rate performance, you will realize actually you are spending more cost to acquire an acquisition on channels that perform at lower conversion rates. When it comes to scrutinizing the conversion rate performance according to the channel you could then refer to your touchpoint attributes and learn what are the winning factors. For example, it could be because channel A is a touchpoint where the ads are being shown to the right audience (audience attribute), with a prominent ads space (medium attribute).

So there you have it! The data-driven online acquisition framework and guide on how to manage your acquisition performance objectively. This article can go on if we also want to deliberate on other reporting models (also known as attribution models) instead of last non-direct click. For instance, you may want to accredit your acquisition evenly across all touchpoints that took place during the purchase journey, instead of only the last clicked touchpoints. But I guess we’ll just save that for another day 🙂

Skye Lee

Skye Lee

Digital Evangelist | Data Driven Marketer | Tech Startup Marketer

Let's get in touch