Guest host Simon Mathonnet from Unbounce sought PPCers views on the techniques or tools they are using for optimization apart from A/B tests, what factors do they take into account while optimizing, what types of campaigns do they generally use landing pages for, and more.
Here is a screencap of this week’s PPCChat session
Q1: Do you have any input or responsibility for landing pages or CRO in your accounts? If so, what is expected of you?
Some, but not much. Mostly I can just point out if there is an issue, and run quality score reports. That’s about it. @JuliaVyse
Yes, we are usually in charge of CRO input on landing pages. @jord_stark
I don’t have full ownership of landing pages, but I do provide feedback to try and improve performance. @FindingAmanda
Depends on the client. Some clients we do all the landing page work. For others none at all. We’re much more involved for lead gen clients. @CJSlattery
Everyone doing PPC should *at least* have input on Landing Pages and CRO. We’ve got the data, to back up our recommendations. @Realicity
I don’t seem to have as much as I used to, which I don’t love. I always review landing pages and those of competitors and have a list of suggested things to do now and to test though! @NeptuneMoon
As a paid media specialist, even if it’s not our “job” to create LPs, it’s ultimately our responsibility to speak up when a landing page isn’t suitable or needs work. We often forget how important the post-click experience is and yet we’re being held to conv. @owlflurry
Sometimes yes, sometimes no. @stevegibsonppc
Spend 2 years at @unbounce and worked on more LPs then any paid marketer in the world. We make the ecom/DTC/SaaS LPs for all our clients. To many come back broken or have a word salad written on them. We dabble in CRO for clients we want a deeper relationship with. @duanebrown
Yes, I am expected to provide guidance and feedback on the landing pages. Also reviewing how people interact with and the results from the landing pages. @lchasse
I try to have that responsibility. I find it very valuable to have a designated place to send traffic and be able to control as much as possible.. CRO is also something I really enjoy – it’s a holistic practice that lights up a lot of my skillsets. @ferkungamaboobo
1. I hold most of the responsibility of landing pages in CRO in my accounts (unless the client has their own design and dev resources). @MarketingByMark
2. I used to be a designer and have an interaction design degree so I take great pride in this. If we’re using @unbounce I take on full design, copy, and CRO (with help from others for builds). If they need to be developed, I take on wireframes, copy, and CRO. @MarketingByMark
Q2: What types of campaigns do you generally use landing pages for? When/would you use an existing page of your website instead?
Use them always for lead gen campaigns. Use them rarely for ecommerce. Most ecommerce stuff is going to website pages, including product pages. @CJSlattery
I use landing pages for all of my lead gen campaigns (unless they’re lead form only). I prefer to use dedicated LPs for all paid campaigns, and absolutely always have customized LPs for paid search. I only use web pages for ecommerce and awareness. @MarketingByMark
Brand campaigns usually use the website. Any lead gen efforts I want going to a landing page. @robert_brady
I prefer distinct landing pages wherever possible for both customization and analytics/tracking. Sometimes though, it just isn’t possible to have separate LPs so you have to make the best of what you’ve got. @NeptuneMoon
Sale pages, lead forms, offers on specific products are good examples of where a specific landing page can be beneficial. For ecom a lot of times, the category pages themselves can outperform specific landing pages, so it is good to test. @lchasse
Not all brands need LPs. Sometimes sending traffic to your site/homepage is the right call. All our SaaS/Subscription clients get an LP. It is a case by case thing for ecom and DTC clients. We try to focus on the biggest traffic pages where we want to test/improve UX @duanebrown
When better copy will significantly move the needle. @stevegibsonppc
For direct lead generation, very specific actions. @andreacruz92
Most. Lead Gen especially. Ecom sometimes. Usually will test LPs vs product category pages and other pages already on site. Then follow the data. Budget can affect LP usage. If lower budget, then might roll with existing page with CRO recos, then view results. @Realicity
For B2B clients I typically use a specific landing page. For eCommerce, I use pages already on-site like collections, best sellers, and sometimes right to the product page. I spend time and energy on selecting the best LPs because I feel it is important! @FindingAmanda
We use a ton of landing pages at Unbounce (duh!) – for most of our campaigns actually. We might rarely drive users to the website instead when our goal is for users to browse the website and engage with any other page instead of taking a single focused action. – Simon @unbounce
It’s worth noting that different content and offer require, or not, unique landing pages. +Newsletter sign-up – dedicated page +Ecom upper funnel – Cat page +Ecom lower funnel – Prod page +Lead Gen – dedicated page. @Realicity
It really depends on the client, the situation + the objective; for introductory campaigns, sometimes a homepage or relevant category page performs better than a LP. In other cases, we’ll use a landing page for a specific service or offering, or product. Depends. @DigitalSamIAm
Depending on the industry and performance. Sometimes I will run an A/B test between the LP and the website, and I’ll run ads to the website if it wins out in the A/B test. @jord_stark
I’d argue that any campaign, regardless of ad channel, needs a dedicated landing page. It’s all about consistency in copy and design, being able to track campaigns, and having a canonical place to have the information for a campaign. @ferkungamaboobo
Q3: What other techniques or tools for optimization, apart from A/B tests, are you currently implementing?
Hands down @hotjar … great tool for running surveys and trying to talk to your customers more. I would take talking to customers over running 12 a/b tests any day of the week and twice on Sunday. @duanebrown
I pull weekly reports on lead progression and meet with sales teams to discuss quality to identify where adjustments can be made. In this vain I also test CTAs to not only impact CVR but also find out WHO they resonate with. @MarketingByMark
I have done feedback sessions with clients/customers to get feedback on everything from creative choices to navigation. We have also done usability studies with specific segments to find opportunities, hurdles, etc…@lchasse
We use @hotjar like a religion; testing out Clarity from MSFT. Google Optimize is fantastic, both for A/B + MV Tests, but also personalization at scale. @DigitalSamIAm
For terms with a lot of search volume: Having the traditional PPC form with a few bullets vs. adding different items to the page, providing more information (video + asset + demo) – to give options to users when intent is not clear and there are many competitors. @andreacruz92
I like to get info from the sales team if possible to figure out where things are not working quite right. Customer reviews can be a goldmine for what to test too! @NeptuneMoon
We’re using our own new AI feature, Smart Traffic for our landing pages. In our latest campaign experiment, we saw a 12% lift in conversion rate compared to an A/B test, while the average lift is 30% across all pages with the feature enabled – Simon @unbounce
@gtmetrix to check page speed (easiest win for CRO) On LP with Dynamic Headlines to keep things concise. @yaelconsulting
Most of the time, A/B testing isn’t doing what you expect it to. Except in very large traffic situations, I’d argue to just make the change and see if it works. @ferkungamaboobo
Q4: What factors do you take into account when prioritizing experimentation and optimization?
Traffic volume. @yaelconsulting
When prioritizing experiments I consider: • Expected impact • Immediate need/how well is performance now • Levers we can pull i.e. do we have multiple offers to test • Time and resources. @MarketingByMark
Volume and Impact Volume can be related to % of Budget, or % of Traffic. Impact usually relates to Best sellers or higher margins. Usually not worth testing on the worst performing 20% of products/services. @Realicity
#1: The likely upside of the test v the setup cost of running the test. #2: What “level” the test is at. Test big elements before you test small elements within the big elements. e.g. Test the headline before you test the opening paragraph. @stevegibsonppc
In general, I think most testing programs are too ad hoc + unstructured; it’s Person X thinks this, let’s test or Exec Y wants this ad but we know that sucks so “let’s test”. I think you need to identify what you’re solving for before you start testing stuff. @DigitalSamIAm
Potential impact. After a certain point, you can be wasting resources testing things for tiny potential returns. @NeptuneMoon
1: There’s also the reality of regression to the mean; when you have a high converting page, the distribution of likely outcomes tends to be left-skewed; often clients get discouraged when something “doesn’t work” + then give up. @DigitalSamIAm
In any case the factor I focus on the most is the clients KPI. Each client has different goals, keeping those in mind helps get the best results. The real question is what factors can be changed in the LP or Opt. that will lead to a better return on the clients KPI. @jord_stark
What will potentially have the biggest impact on the bottom line. Radical changes are better then doing micro tweaks on an LP. @duanebrown
Conversion elements first – for lead generation, a mini form & a long form, and for phone-based businesses their phone number. This is usually a design focus. The next would be copy – I’m a huge fan of longer, more in-depth copy on landing pages. @ferkungamaboobo
Q5: What’s your biggest hurdle when coming up with new testing or optimization strategies?
Resources and Return – There is an expectation that every test will improve results by X%. We all know that’s not the case. So when we hit those tests where the return was not there, it becomes difficult to maintain the resources for further tests. @Realicity
I think it’s probably one of the following: (1) getting people to see the value of testing stuff; (2) moving people from “testing = good” to “strategic testing = better” or (3) getting execs to live by the data-informed mantra they spout. @DigitalSamIAm
Usually it’s client resources. @jord_stark
Time + traffic. Not all clients have tons of traffic. Some are working on a tight deadline. They need to see x results by a certain date. @duanebrown
Volume of data! A lot of lower budget advertisers have low budgets, which makes it difficult to get enough volume to make statistically significant choices. @amaliaefowler
Technology – the client’s website, what they CAN do. @yaelconsulting
Honestly time and platforms that would make this process easie. @andreacruz92
Inertia. Or not wanting to actually test things that could be significant. Like the offer, the imagery, the headline… @NeptuneMoon
Hello #ppcchat fam. I’m late but I’m here! I would say my biggest hurdle are 1. Finding a strategy that can yield statically significant results, not directional 2. Time. It can be difficult to justify the hours to clients unless performance is catastrophically down. @MarketingByMark
For us, it’s more about prioritization than coming up with new ideas – where we want to run multiple experiments across the funnel while maintaining data integrity and not have cohorts conflict with each other, skewing our results – Simon @unbounce
Q6: Do you have an example of a time when you made a very interesting discovery based on a website or landing page experiment? How did you leverage that discovery?
Images of happy people far outperform unhappy people. Client provided a solution for a problem, but all of their print ads showed people upset or stressed about said problem. I begged for happy “we solved their problem” imagery and it performed WAY better. @NeptuneMoon
2020 has not been the best year for testing for us. Many resources have been reallocated to other areas this year. I learn a ton from other sharing their tests like the ones at @GuessTheTest @Realicity
1. For a data visualization SDK client I’m testing a demo vs an SDK trial download on paid LPs. I assumed that the SDK would convert higher than the demo (which is true), but I found that the people converting on each are completely different. @MarketingByMark
2. So now I’m auditing the account to figure out how I can make sure that demo oriented people are seeing a demo and SDK oriented people are seeing an SDK download. A good reminder to look past CVR and also consider WHO is converting. @MarketingByMark
3. Another test that surprised me was form at the top of the page vs the bottom. In my B2B ppc career I was always taught to put the form at the top for higher CVR, but I now test this for many clients. @MarketingByMark
4. I have found that form at the top generally works well for straight-forward B2B categories, but for newer markets prospects require education and may be more likely to convert if you allow them to digest the page content before presenting a form. @MarketingByMark
Long-form pages significantly outperforming short-form pages after finally convincing a client to test something outside the box of their very-minimal-LP setup. @timothyjjensen
We tested ‘Due Today: $0’ in our signup flow for our free trial and it performed slightly worse than showing the full price due after trial! It prompted us to look more holistically at how users interact with and perceive price at this stage of the funnel – Simon @unbounce
Q7: Do you have questions for @SimonMathonnet & @unbounce ?
Any plans to start offering fully responsive templates for Unbounce pages (if this doesn’t already exist and I’m not aware)? @timothyjjensen
We have a couple things on the roadmap that we think you’ll like! Stay tuned. @unbounce
Where do you see the future of landing pages going? @NeptuneMoon
I believe AI will be a big component of the future of landing page creation and optimization. Leveraging AI insights and tools along with our skills to drive value through landing pages is becoming increasingly accessible, and I have to say it’s very exciting! – Simon @unbounce
Is Unbounce working on stronger individual modules for landing pages? I think the entire page templates are great, but I prefer to build my own design from scratch and would love an easier way to drag and drop customizable modules. @MarketingByMark
Oh do we ever. Happy to pop some case studies/examples over your way: @unbounce
2. Do you have any success stories for smart traffic on @unbounce? I would love to test but “smart” marketing tools make me so uneasy these days. @MarketingByMark
Do you think that @unbounce will ever offer LP building as a streamlined service or help connect customers with unbounce experts? @yaelconsulting
PPCChat Participants
- Julia Vyse @JuliaVyse
- Jordan Stark @jord_stark
- Amanda the Elf! What’s your favorite color? @FindingAmanda
- Collin Slattery @CJSlattery
- James Svodoba @Realicity
- Julie F Bacchini @NeptuneMoon
- Jennie @owlflurry
- Steve Gibson @stevegibsonppc
- Duane Brown @duanebrown
- lchasse @lchasse
- Doug R Thomas, Esq. @ferkungamaboobo
- Mark T Saltarelli @MarketingByMark
- Unbounce @unbounce
- Robert Brady @robert_brady
- Andrea Cruz @andreacruz92
- Sam @DigitalSamIAm
- Lior Krolewicz @yaelconsulting
- Amalia Fowler @amaliaefowler
- Tim Jensen @timothyjjensen
- Simon Mathonnet @SimonMathonnet
Related Links
Stop the wasted ad spend. Get more conversions from the same ad budget.
Our customers save over $16 Million per year on Google and Amazon Ads.