This week's episode is going to tackle a great question that was sent in by a fellow copywriter:
How can I use tools like Google Analytics to diagnose the problems a web page is suffering from?
Thanks to Marcus for sending this one in, because this is an ongoing challenge for SO many copywriters looking to improve their game.
Who this lesson is really for
Content might be king, but let's be real:
Most copywriters want to be more than just wordsmiths — they want to able to give clients surgical, data-driven, high-impact recommendations for improving a funnel's performance, but ...
... they (understandably) feel nervous about stepping off the safe, familiar shores of writing words and wading into the bottomless, monster-filled ocean of crunching numbers.
Even copywriters who muster up the courage to enroll in an analytics course end up quitting because these courses are tailored to the needs of in-house marketers and analysts, a.k.a. people who need to know how to configure and maintain a whole host of dashboards & reports that go way beyond assessing specific copy performance.
What a copywriter needs is enough know-how to quickly dip their toe into a client's existing analytics platform, assess how visitors are interacting with key page(s), and GTFO without getting sucked out to sea by a rip tide of technical configuration problems.
Which is exactly what I'm going to show you!
Over the course of the next few posts, I'm going to show you:
- How to quickly assess a prospective client's analytics / GA setup before even diving into their analytics backend;
- How to check for the biggest Google Analytics problems that can invalidate a client's conversion reports;
- How to zero in on the traffic that's hitting your target landing page or sales page, and
- How to identify where the biggest traffic dropoffs are happening upstream, downstream, and on a given page in Google Analytics, and how to interpret what you're seeing.
"Wait, wait, wait ... why is this important for me to know as a copywriter, again?"
Think of it this way:
Imagine you were hired to help increase the sales in a REAL store, pre-Internet.
Would you make changes to the signs, shelves, and product labels inside the store without watching how customers enter, walk through the aisles, and pay for the product?
Of course you'd go in the store to see what's actually going on first.
By getting a quantified, full-funnel view of how real prospects are interacting with the page, you'll be able to make smarter, more specific improvements that have much higher chance of having a direct, measurable positive impact on your client's conversions.
Today's lesson: How to sanity-check your client's analytics before diving into any data
Happily, this part does not actually require any technical skills — all you need to do is ask your client the right questions, which you can ask during your consultation / lead-qualification call.
Question 1: "What's the #1 MACRO-conversion goal for the copy you want me to write?"
Notice I don't just ask "what's your conversion rate?" because even the term "conversion rate" can be interpreted differently from client to client.
By "macro-conversion," I specifically mean what is the ultimate, business-critical end-goal we want users to achieve after landing on and consuming the page copy?
Macroconversions are those bottom-of-the-funnel transactions that directly impact the company's revenue. 99.999999% of them involve the user filling out and submitting some kind of form.
- Sales/purchases (i.e. one-off payment is transacted)
- Paid subscriptions (i.e. a free-trial user upgrades to a paid plan)
- Sales leads (i.e. contact information is given with the intent to negotiate a sale/purchase)
- Trial signups (i.e. user creates a free account for limited use of a product)
- Mailing list leads (i.e. contact information is given in exchange for some kind of free offer/value.)
Notice that with the exception of #5, macro-conversion goals often do NOT actually occur on the page you're writing — they occur one or multiple pages/clicks downstream of the page you've been assigned.
Question 2: "What is the #1 MICRO-conversion goal for this page?"
Because the macroconversion goal can be several pages further into the funnel than the sales page you're optimizing, it's equally important to know what the #1 micro-conversion goal is ON the page itself.
In other words, what do you want the user to do (or where do you want them to navigate to) immediately after they read your copy?
The most common example of a microconversion goal is clicking a call-to-action button or link that leads to the next page in the funnel.
(You'd be amazed how often the primary microconversion goal for a page ISN'T clear. Very often pages will have a dozen links and buttons on it, which in itself can have a negative impact on your conversions.)
Once I have the top microconversion goal of a given page and the macroconversion of its downstream funnel, I officially have a defined user flow to assess and start optimizing.
... Now we're getting somewhere!
Question 3: "Where is this funnel's traffic coming from?"
This question is critical because it gives me an early sense of the mindset and awareness level of the page's audience, and in many cases what visitors' expectations will be before they read a word of my page copy.
I can't stress this enough:
Being aware of what people searched for, read, and clicked on immediately before entering the funnel you're optimizing is one of THE EASIEST AND MOST CONSISTENTLY IMPACTFUL weapons you have as a digital copywriter, because it allows you to "message-match" your opening copy to precisely what got the visitors' attention and compelled them to seek you out in the first place.
If most of the page's traffic is coming from SEM ads: You're going to need to find out what the top-performing ads are, and make sure your headlines line up with their copy.
If most of the page's traffic is coming from organic traffic: You're going to need to find out what the top search keywords are, as well as the associated SERP copy, and align your page's messaging to reflect them accordingly.
If most of the page's traffic is coming from BRANDED organic traffic: You're going to need to adjust your copy to speak to an audience with a higher level of product awareness (because they already searched for the product name). The value proposition on the page will likely have to be more feature-specific and less general problem-focused than it would be if you were targeting prospecting traffic.
If most of the page's traffic is coming from referral websites: You need to find out what the CTA text is that people clicked on on those websites and make sure that what they see when they land is related to that text.
|Primary traffic type||Upstream copy you should to "message-match" your opening copy with|
|SEM / PPC ads||Top-performing ads (especially their headlines). In the case of things like Facebook ads, your imagery & branding should also match up, if it's applicable and possible.|
|Organic traffic||Top search keywords and/or SERP copy, within reason. Don't over-search-optimize your headline to the point that they lose emotional appeal or are less clear at a glance.|
|Branded organic traffic||Adjust your copy to speak to an audience with a higher level of product awareness (because they already searched for the product name). The value proposition on the page will likely have to be more aligned with specific, unique outcomes and product features.|
|Referral traffic||The CTA text that people clicked on on those referral webpages.|
This isn't to say that the opening copy of your page has to exactly match the copy of its traffic sources — it doesn't — but making sure that it at least is clearly guiding the audience toward the same general goal will prevent a significant proportion of visitors from bouncing.
Question 4: "How many visitors are coming to this page per month? What % of them fully convert (i.e. complete the macroconversion goal)?"
This question helps me get a sense of a) the volume of traffic hitting the page I'm being asked to optimize and b) what proportion of this traffic is reacting well vs. poorly with the existing copy.
More importantly, this question (as well as #3 above) gives me a sense of how familiar this company is with their own analytics.
If they can immediately tell me the amount of traffic hitting the page (not the entire website) as well as its conversion rate, that's a good sign.
If they waffle and say "hmm, we'll need to get back to you on that" this might indicate that their analytics setup isn't being regularly maintained.
Which leads us to the next & LAST question ...
Question 5: "What tool(s) are you using to measure the performance of this page?"
At minimum, the client should have the following installed on their website:
- One robust analytics platform (Google Analytics, Adobe Analytics, Heap, Mixpanel, etc.)
- One heatmapping & visitor-recording app (e.g. Hotjar, Clicktale, LuckyOrange, etc.)
Sometimes your client contact won't even be sure of the analytics they have set up, because it's not their department and they rarely ever talk to the developers who are in charge of it.
In this case, you can quickly doublecheck what's installed on the page yourself using a free browser extension called "BuiltWith."
When I scroll through the BW breakdown of a given website, at minimum I want to see Google Analytics & Google Tag Manager listed in the 'Analytics' section.
If the client has GA installed and it's been installed correctly, I'll at least have pageview data to dig into, even if they haven't been actively managing the account.
Sometimes you'll see more than one analytics platform installed (I've seen some websites with Google Analytics, Mixpanel, Heap, Kissmetrics, all installed simultaneously), in which case the client is going to have to clarify which tool they're actually using to measure the performance of the funnel.
Unfortunately, what BuiltWith won't tell you is whether or not Google Analytics is installed correctly, but thankfully tracking-code detection tools like gachecker.com can help with this.
GAchecker works by crawling a website's page code and checking to see which Google installation codes have been hardcoded into the site's pages.
These days it can be trickier to interpret GAchecker than it used to be because hardcoding analytics tags into a website is becoming less and less common, but it's still very useful as a sanity-check before diving into a client's GA account.
Watch the videos below to see 3 examples of different ways you may see GA show up in BuiltWith and GAchecker and what you can infer about the quality & cleanliness of the client's GA data from what you see with these tools.
Bad Example: Ontario government website
TL;DW: Builtwith shows GAnalytics is there, but gachecker is showing that GTM and GA are hardcoded into the page => this is a red flag, because they didn't actually install GA properly, and very likely have duplicate hits firing and being sent to GA.
Good Example A: Copyblogger.com
TL;DW: Builtwith shows many Googla Analytics tools installed, but gachecker only shows Google Tag Manager. This is actually a good thing, and we can infer that GA has probably been installed properly.
Good Example B: Copyhackers.com
TL;DW: Again, Builtwith shows that multiple Google Analytics have been installed, but gachecker doesn't show ANY. In this case, look for a tool called 'Segment.' If it's there, this indicates there is a dedicated person overseeing the analytics, but they're doing it through Segment, rather than Google Tag Manager or something else.
To sum up, if the client's analytics setup is showing something weird — like the first example, where GAchecker showed duplicate or conflicting GA installation tags hardcoded into the page — this is a red flag in terms of the validity of their conversion data.
It doesn't necessarily mean I won't take on the project, but I will let the client know that if their analytics aren't set up properly, I have less to work with and won't be able to provide as many diagnostic insights as I normally would.
(Moreover I would stress that this is an issue they really need to escalate and have dealt with if they want to do data-driven optimization in general.)
If a client's analytics setup falls under "Good Example A" or "Good Example B" as above, I'm generally pretty confident that GA has at least been installed correctly, and I will move on to checking the cleanliness of the analytics data in the company's GA dashboards.
At this point, without looking at a single byte of analytics data, I've got a defined, clearly bounded problem I can start optimizing for this client.
By asking the client just a few questions, I now know:
- The business-critical goal at the bottom of the funnel,
- The initial entry-point goal at the top of the funnel,
- How people are getting to the funnel and from where,
- How many of them are making it through to the end.
What I DON'T know is what's actually going on INSIDE this funnel ... and that's what we'll dive into next time.