• Bryan Ong

5 Lessons I learnt from directly managing millions of dollars of digital advertising spend --- Ep. 1

Updated: May 7, 2020

As a marketer specialised in digital advertising, I have spent more than 4 years implementing campaigns and measuring performance. During this time, I spent slightly over 3 years in my career as an in-house specialist and am now a founder of a digital marketing agency, Method B.

Through a combination of my career and now my agency I have directly managed and implemented millions of dollars of digital advertising spend. The purpose of this entry is to share with you what I have learnt from this journey so far, though I am still learning everyday.

While digital marketing is a hot topic right now, I believe that there’s as many of you reading this who are an expert in this topic and as many of you who are fairly new to this. Thus, I will do my best to provide my learnings in a mix of both at the strategic as well as at a technical level. As you have seen from the title this is the part 1 of several parts (I am deciding how much more content this should be) and as recommended by my wife as well who told me its more like a serial drama than a boring old listicle.

And first, a disclaimer: due to confidentiality and to protect clients’ information, no particular client or industry will be mentioned and the learnings provided are by no means a guaranteed way to help you achieve better campaign performance due to multivariate nuances and factors.

Now that we have gotten these out of the way, let us continue.



1) Different platforms serves different purposes

The top digital advertising platforms (excluding those in China and Korea) are still these top players: Google/ YouTube, Facebook/ Instagram and LinkedIn.

Lets explore each of these channels and what they can be used for, note that these are not exhaustive and meant as a quick way to understand what each platform can do. There are lots, lots more to what these platforms can do, but I will just list some of them.

So what: Knowing each of these will allow you to make better decisions on your digital advertising strategy for your company/ clients or know what to ask of from your agency should you be engaging one.

a) Google Search Advertising

  • Enables you to rank above other organic listings in Google Search based on your bids

  • Feature limited-time only promotions on your ad copies

  • Captures high quality, high intent searches through Google during purchase consideration

  • Direct targeting to customers based on their location when they search

  • Bid on specific keywords and keyword variations, so as to capture demand based on what they are searching for

  • Negative keyword targeting

  • Make use Google’s algorithm to automate bidding, maximising towards specific goals

  • Retarget customers who visited your company’s website

  • Set audience targeting to observe which are the top performing demography/ in-market/ affinity groups of audiences and adjust bids accordingly

  • Responsive Search Ads (RSAs): enables you to feed a set of headlines and descriptions for your ad copies into the platform to churn out the best performing combination of text


b) Google Display Network Advertising

  • Great way of retargeting customers who visted your website and to continue showing your advertisement to them outside of Google search

  • Direct targeting to customers based on their location when they search

  • Set audience targeting to observe which are the top performing demography/ in-market/ affinity groups of audiences

  • Responsive Display Ads (RDAs): feed a set of headlines, descriptions, images and logos for your display banners and allow machine learning to find the best performing combination of text and images

  • Exclude specific websites or categories to not show your display banners on

  • Implement Google Shopping campaigns that links your inventory catalogue with third-party platforms like Shopify. This allows you to show selected products from your e-commerce store to customers through the Google Display Network.


c) YouTube Advertising


d) Facebook & Instagram Advertising

  • For video and image push advertising, featuring a good suite of targeting options

  • Direct targeting to customers based on selected demographic options, interests and locations

  • Great for growing new customers and market share

  • Features powerful targeting algorithm and lookalike audiences (other potential customers who look like your current customers)

  • Remarket to customers who visited your websites before

  • Allows for video sequencing based on % of which videos watched OR number of days they been to your website or added to cart

  • Specifically useful to target add to cart drop-offs customers with promotions

  • Allows for boosting of organic posts from Facebook and Instagram

  • Popular placements such as feeds in carousel and single video/image formats, as well as stories.

  • Direct advertisements to Messenger for customers to chat directly with businesses


e) LinkedIn Advertising

  • Specialised in a formalised, business setting, with strong intentions of B2B targeting

  • Features video and image push advertising

  • Target customers based on skills, seniority, job functions and company size, among many others, which are uniquely effective to LinkedIn

  • Retargets customers who visited your website

  • Great for building strong B2B brand awareness and thought leadership

  • Lead generation forms that leads to whitepaper/ analysis report downloads; they are amazing in generating leads


Phew, that’s a really long list, and I tried really hard to keep it as short as I could because there’s just so much to offer in each one of them. Do note that I am not going into the cost per clicks or the cost nuances of each platform because they vary greatly based on keywords, audience targeting, placements and locations. With these in mind, you are now one step closer to creating your multi-channel digital advertising strategy.

2) What should my recommended budget be?

This is a question I get asked very often since I started Method B. Your recommended budget are based on multiple factors such as locations, nature of product, price of product, minimum number of clicks you can get per month and so on. It also depends on your company’s or clients’ budget and so there’s no easy answer to this.

But… one day, should you find yourself caught in a situation when you need to answer this question in 5 seconds and the client is a SME:

  • Google Search Advertising $1000 to $3000/ month

  • YouTube Advertising $1000/ month

  • Facebook & Instagram Advertising $2000/ month

  • LinkedIn Advertising (more for select B2B) $2000/ month


Feel free to recommend 1 to 2 platforms and not all at one go as your company or client may need time to prepare the creatives. Also feel free to shift budget around between platforms based on the clients’ appetite to scale their campaigns.

Essentially, a recommended budget should take into account how much a click would cost, and how many clicks would it take for it to become a purchase or lead, and how much does that purchase or lead mean in dollars to you. Should you still be able to make a profit after deducting necessary costs, then okay, it makes sense. There’s this core performance metric called Return on Advertising Spend (ROAS) which measures how much revenue you generate per dollar of advertising dollar spent. Should the number be larger than 1, then that’s meeting the bare minimum (unless in the rare scenario that you are in a cash burning spree to gain market share), but the more normal number should still be about 2 or more, conservatively. Depending on industry, this can vary extremely widely.

I understand that it can be unreasonable to know what’s the cost per click or cost per purchase even when they yet to run any campaigns before, and so in this case, without any prior data to reference, you may choose to implement a comfortable budget to try out to gather some data, so that you can make a more informed decision of your subsequent budget recommendations.

Also, I don’t believe that anyone can just guarantee ‘x’ amount of revenue or leads using an allocated budget even with industry benchmarks, just simply because every company is different.

We have now covered what is a normal advertising spend for a SME, using ROAS as an indicator of an advertising budget that makes sense and that there should be no performance guarantees in campaigns, especially brand new ones.


3) Finding the Sweet Spot

The best part of having your campaigns run in multiple platforms is that after 2 weeks to 1 month of running your campaign, you would know which platform is the one that is generating the most leads or revenue for you. Using this simple analysis, you can then shift more money from the lower performing platform to the better performing platform.

Usually in a digital campaign in each platform, it is always good practice to set up a number of varying targeting options such as keyword groups, target audience groups or varying creatives. Doing so allows you to find out which groups are the revenue generators while which groups are the cash burners.

One important metric to look at is Cost Per Acquisition (CPA, or also known as Cost Per Purchase or Cost Per Lead CPL), depending on which goal you are optimising towards. Find the ad group with these 3 factors and you would have found gold:

  1. Lower CPA compared to other campaigns or ad groups

  2. Bringing in substantial (30% and above contribution) revenue or number of leads

  3. Spending up to budget limit that you set


With these 3 factors in place, it means that you found a revenue generating ad group, with the ability to scale with additional spend. Now, you can double down on this sweet spot and “exploit it”.

To get the money to double down on, simply move the budget from other campaigns or ad groups which aren’t generating as much revenue.

With these, you can now locate the campaign sweet spot and double down on it, and improve overall efficiency/ performance in all your campaigns.


4) Machine Learning? Algorithms?

These are gigantic words that are used so often these days… but what do they mean for us digital marketers?

Firstly, we need to understand in what form do machine learning manifest in for each platform: they are generally campaign objectives, automated bidding strategies and lookalike audiences, in different forms.

  • Google Search, Google Display Network and YouTube: campaign objectives you select based on customer behaviour e.g. for clicks, for video views, automated bidding strategies towards goals e.g. maximise clicks, maximise conversions, Target ROAS, interest group targeting

  • Facebook & Instagram — objectives you select based on customer behaviour e.g. for Messenger chat, for brand awareness, for lead form fills, for purchases, automated bidding strategies e.g. target cost, lowest cost, cost cap, interest group targeting, lookalike audiences targeting

  • LinkedIn — objectives you select based on customer behaviour e.g. brand awareness, website visits, website conversions, lead generation, video views, automated bidding strategies e.g. maximise for impressions, maximise for clicks, maximise for conversions and lookalike audiences targeting


With these plethora of options in each platform, it can get overwhelming. But fret now, should you be new to this, just select what makes sense to the campaign and let the algorithm take care of the rest. For example, you want to get customers to purchase your product, simply select maximise conversions (do note that you need to set purchases as a conversion goal in the platform).

Or you can choose to take control of some parts of the algorithm, say, Google Search for example, you can choose to override the automated bidding strategy by bidding manually (though this is recommended for advanced digital advertisers only). Facebook allows some form of control over the algorithm as well, for instance, by allowing advertisers to enter a cost cap, which means to get the most possible conversion actions within the allocated cost per action that the live auction bidding permits.

Another powerful machine learning tool that is specific to Facebook & Instagram is lookalike audiences. From my experience so far, they are often times the top generating targeting ad set for scaling with efficiency. You can base lookalike audiences from a few options, for example:

  • From your existing customer base

  • From specific actions that your customers did, such as enquiry submission or add to carts

  • From Facebook & Instagram page engagements

  • From people who watched x% of your selected videos

  • From top x% of people who spent the most time on your website


With the above targeting options, you can set parameters to the machine learning algorithm such as in the past 30 days or the degree of which the find lookalike audience with 1% being the most similar and 10% being the least similar (think of it like a radius with the most inner circle who are most similar).

Machine learning and algorithms doesn’t seem so much of a stranger now is it? You have now learnt how to identify which parts of the platform are using machine learning and also learnt how to take manual control of some aspects of each platform’s algorithm to customise it into one that suits your needs.


5) Creatives are very important

From my experience working with clients and companies, I would say that not all recognise how important creatives are. The truth is that after running the same videos or images for the past 1 to 2 months, we start to experience ad fatigue: click-through rates, conversion rates plummet and cost per click rises.

Creatives are perhaps the 1 largest influencer to campaign performance besides targeting options.

There are a few questions we should consider when coming up with creatives

  • Are there any seasonality factors? Chinese New Year, Valentine’s Day, Black Fridays/ Cyber Mondays (BFCM), Christmas

  • Are you launching a new product?

  • Are you launching into a new market?

  • Are there bundles or promotions?

  • Do you specifically need to build social proofing?

  • Are you building a particular brand image?

  • What are your call-to-actions?

  • Are your videos succinct enough?

  • Are there too much text on your image?

  • Are you trying to tell a story?

  • Are you making use of sequencing to nurture or funnel through your customers?


While there are lots of placements and digital media formats such as single image, single video, carousel images, carousel videos, stories, responsive display banner and more, to choose from, generally, creatives need to have the following:

  • Strong indication of your brand at the first 1 second or on the image

  • For videos on feeds, make it 10 to 15 seconds max

  • Each frame needs to be fast-moving to capture attention

  • Colours tones and fonts need to be on-brand

  • Indicate more brand/ product awareness at the “first-touch” (upper funnel)

  • Go more into product features and social proofing at the “second-touch” (middle or lower funnel)

  • Then strong call-to-action at the “third-touch” (lower funnel)

  • Combine these, measure, and reiterate better


You can come up with the perfect formula to your creatives by combining multiple elements to see which works best.



Conclusion

Thanks for having read this far into the learnings. I hope this helps you gain a better understanding of each of the platforms and the multiple elements you can utilise to create the best digital advertising strategy for you. That’s all for this episode, there are lots more where this came from, so stay tuned.

See you soon in the next one.

Find this useful? Support by sharing this with your network and friends!



About Bryan Ong

Bryan is the founder of Method B, a digital advertising agency which has served clients such as Google, WeWork, Greendot among many others. Bryan is known for his strong work ethics, sincerity and a 360 view of customer acquisition growth, which has led to strong client recommendations and a rapidly growing client base.

As a pure digital advertising marketer, throughout his career, Bryan has directly managed millions of dollars of advertising spend across multiple industry verticals in APAC and has developed systems that has consistently allowed him to perform above industry benchmarks. His expertise includes strong experience in APAC markets, launching new brands and targeting premium customer segments.

In his earlier career, Bryan was also a TEDx speaker, a motivational speaker whose books sold in Kinokuniya and Popular bookstores and a SPRING Singapore (currently known as Enterprise Singapore) Executive Development Scholar. He was featured on The Straits Times, Channel U, Capital958 FM, Lianhe Zaobao, RollingStone, e27 and other publications. Enquiries and collaborations? Email him at bryan@methodb.co (not .com)



Footnotes: [1]: Paige Cooper. (17 December 2019) 23 YouTube Statistics that Matter to Marketers in 2020 https://blog.hootsuite.com/youtube-stats-marketers/#user [2]: Think with Google (October 2019) Planning for 2020? The 5 rules of engagement to deliver results on YouTube all year round Google/Ipsos Lab Experiment, U.K., n=1,000 people age 18–64, April 2019. https://www.thinkwithgoogle.com/intl/en-gb/advertising-channels/video/rules-engagement-deliver-results-on-youtube-all-year-round/


0 comments