Privacy vs Tracking – Round 10: Fight
The solution to iOS 14+ privacy settings is already available – server side tracking.
Not widely adopted (extra costs for server and implementation), but I am sure it will be within the next year…. (unlike full GA4 adoption, which is still a while away and only worthwhile for specific use cases i.e. data warehousing/BigQuery integration and predictive analysis)… BUT if advanced segmentation and behavioural targeting is part of the paid strategy, then server side tracking should be implemented by default.
I have noticed that it’s become a bit of a buzz word along with ‘Facebook API’ with the PPC Agencies sales pitch, whether or not they understand the intricacies of it. So I hope to shed some light on this topic…
My observation (from the 50+ GA accounts of $10m+/year ecommerce sites that I have had access to in the last year or so), is that the majority of paid campaigns are not doing any kind of advanced targeting anyway (e.g. segments of visitors with a high probability to convert)… most are blanket campaigns for brand or product terms – which is good for overall brand awareness and down the funnel sales, but profitability/ROAS can not be optimised much on that…. and in these cases it’s only going to affect ad platform attribution data such as revenue/CR%…. not the bottom line business sales/P&L reports in the CMS/ transactional platform. So it’s not going to result in a decline in overall profitability.
Post Covid World
Advanced targeting goes hand in hand with conversion optimisation. If you took a look at ecommerce businesses doing $50m+/year complaining about declining performance you’d find unoptimised processes and untargetted campaigns – no doubt……….. BUT I also have a post covid opinion:
Covid was significantly great for ecommerce businesses. It presented a period of massive growth for many and a nursery ground for many emerging players establishing themselves. People were shopping online more than ever. Shopify stock price grew from 325pts (March 2020) to 1650pts (July 2021) – 407.692% increase. No, I didn’t buy their stock, but I wish I did (don’t buy them now).
The bottom line is the shops are open now.
Post covid, that is bound to drop. People spent enough…. and the spenders in the economy are going to be spending on entertainment next – clubs, events, restaurants, holidays, etc.
So a decline in retail ecommerce profitability is inevitable. This needs to be accepted in a business sense, and profits re-invested into customer retention and sustainable growth in the long term…. not chasing peak sale periods in the short term. The last year+ was a unique situation.
How Server Side Tracking Works
Anyway, back to the iOS14+ issue and SS tracking – in simple terms this means:
- Server to Server tracking/data transference a dedicated tracking server to the FB or Google Ad platform server.
vs…….. what is currently happening, which is:
- Browser to Server (safari, chrome, etc. to fb or google server). It’s the browsers that are filtering and re-writing common tracking scripts such as FB and GA.
Analytics in a GDPR Compliant World
If you were not already aware, implementing server side tracking can be a grey area in terms of legality, because it by-passes consent and by-passes the new in-built privacy/protective features in browsers (depending on how you set it up).
Server side tracking = no restrictions are automatically enforced. BUT Shopify already offers the FB API (server side) within their native FB app. Enough said.
Costs vs Benefits
Server side is new-ish to GTM (free)…. but has been a long standing feature in other tag management platforms such as Tealeum and Adobe (paid) – at the end of the day you get what you pay for …but as I said I think this matters more if advanced segmentation and targeting is the primary strategy for re-marketing audiences.
If you are implementing a server side tracking implementation via Google Tag Manager, all you need to front is the per usage hosting/bandwidth costs. Making it very affordable as this can be pitted against campaign spend.
Tealium is around the mark of $150 – $450 / month and Adobe is significantly more.
For blanket campaigns (brand/non-brand keywords) attribution data may be missing, but business profitability won’t be affected. If there is a decline it is for other reasons. So there is no real need to spend on server side tracking, unless it’s just to asses/justify ad spend.
The Future of Supply Chain Management
I have seen many complaints of supply chain issues causing profitability issues and comments around this instigating a migration of manufacturing to home soil (US or EU) in the next decade – my opinion is:
Nonsense. No way.
Not unless end-product prices become astronomically high.
Firstly, we still need to source raw material (Cobalt, Silicone, Copper and Quartz for chipsets in any electronics from China, Brazil and Africa) and the majority of the worlds cotton comes from China, India, US and Brazil…and then secondly the cost of the labour force is not going to be cheaper than shipping costs.
Why move production to a more expensive labour force if highly skilled Western workers are not needed? Machine lead/non-manned factories would need to become dominant.
If profitability is principal over quality (which it has been on the mass scale historically), then larger brands are not going to do this…. only the small to medium enterprises could consider it. Small are probs already home soil… but their profitability is on the lower end of the scale, and they will want to grow eventually.
Saying that the world is changing, the income gap is becoming smaller… but the West still houses the most skilled workers which are dedicated to high-tech, creative and finance industries and developing nations are more suited to manufacturing and a source of affordable craftsmanship. Supply chain/shipping process evolving into more refined practices out of the current deficit is much more likely.
The Key to Profitability
If I were to survey 50 $50m+/year ecomm businesses I would find out what their most profitable products were and find out why, then what their highest converting acquisition sources are, then highest performing campaigns (i.e. brand vs product vs generic vs remarketing) …and then scale and cut accordingly. The method is simple enough.
…and if I wanted to increase Revenue and CR in the short term it would all be focused on technical on site analysis to identify and fix UX fails. Mid to long term – not increasing ad spend, but optimising return on ad spend by behavioural retargeting.
…and, yes. I would push them to implement server side tracking much sooner than suggesting GA4, but for data analysis and to enable profitability in advanced remarketing campaigns… not just to tick a box on a list of services offered.