Have you ever pulled up a website on your mobile device only to find that there are so many buttons and conversion widgets on a site that it begins to actually cover up the content of the webpage? Have you had a pop-up that’s so aggressively big that you could not click out of it? Have you experienced two chat boxes appearing with a third chat button on the side of the screen? Perhaps you should open your organization’s site and see if it happens to you. It may not be the only thing impacting your website conversion rate.
What started as some friendly banter on LinkedIn became a research project that took more time than needed, but the work begged to be done nonetheless. The post’s premise is that “VDP Vomit,” or all of the buttons, widgets, and distractions, also known as Calls to Actions or Calls to Action (CTAs), limit vehicle detail page conversion. The link to the post can be found here; however, it may not be accessible due to LinkedIn’s policies. The post’s author summed it up by saying, “By decluttering, you transform your site from a maze into a journey.” In other words, please keep it simple for the person you’re trying to convert into a customer.
In the ye olden days of web design, designers didn’t take long to figure out that a few giant buttons meant better results. That’s why it shouldn’t be a surprise that nearly every major transactional site that doesn’t belong to a dealership keeps the call to actions to a minimum. If Amazon can process 12 million transactions daily with just a couple of CTAs above the fold, why can’t automotive dealerships do it, too? Carvana moved over 200,000 units in six months with just one button. Tesla did over 600,000 units in 2024 doing the same thing. Maybe they’re onto something. Maybe it’s not that easy.
Setting the Level
Let’s clear some assumptions on both sides of the argument. First, not all dealerships are created equal (which is why Tesla and Carvana do everything possible to eliminate the franchise bottleneck). Second, the ability to hire/train/retain the proper staff limits throughput and consistency at any given dealership (which is why Amazon is struggling). Third, forced OEM website requirements and OEM lead volume often put a crushing burden on certain franchises, further restricting the consistency of the staff and, thereby, the transactional results. Fourth, there are no measurement standards for retail automotive KPIs, so it’s easy to challenge results without showing any math. Fifth, there is subjectivity regarding what metrics matter, with a strong bias towards volume. And finally, there is a notion that organic traffic is not limited by seasonal fluctuations and macroeconomic norms, but is limitlessly free. This is much more complicated than a &$#@ measuring contest.
The Case for Fewer CTAs
There is plenty of academic research that supports fewer CTAs. I recommend that the curious use Perplexity to research since Google makes the majority of its money from digital advertising. Specifically, I would refer to the Hick-Hyman Law/Hick’s Law, decision fatigue, or what’s colloquially known as “analysis paralysis.” In a nutshell, someone will choose from fewer options more quickly, while additional options require continuously more mental energy to process. If you’re a frequent grocery shopper, this is something you experience on the regular.
This phenomenon is not exclusive to websites. Reducing decisions increases the velocity of a positive outcome, whether it be landing on the right vehicle, setting an appointment, or closing a sale. When given the option, DealerKnows has routinely increased year-over-year results when implementing Alternative Choice Closing and Assumptive Closing techniques into inbound and outbound communications. Further studies in behavioral economics have widely proven reducing decision fatigue is a sound principle, and if your website provider allows for it, it’s something you can easily put to the challenge through A/B testing (but, just test for one variable at a time).
Still, inquiring minds need to know. That’s why AI pioneers Vast/CarStory (acquired by Vroom) researched the number and syntax of calls to action to answer the question. Vast’s team took the research down to the semantics of the words on the buttons, even the color, size, and shape. Its research concluded that many CTAs have little to no value (published in a whitepaper, How to Convert Used Car Shoppers to Buyers, March 2015) when generating lead submissions. At the time, Vast’s team comprised many brilliant minds, including those with PhDs in Computer Science and Cognitive Psychology, as well as those who invented the very backbone of automotive e-commerce. The standards used exceed anything in automotive before or since. The team concluded that dealerships should not overwhelm users with too many CTAs and should focus on the ones that will move buyers through the process. Case closed.
But…
What if traffic isn’t your goal? If you find yourself asking that question, give yourself a gold star. Depending on where you sit on the side of websites, your goals may be very different. If generating the most lead submissions possible from a website is what puts food on your table, then how those leads convert into income for the client isn’t a problem until they bring it up (if you want to test this, threaten to cancel your third-party lead provider and watch your lead quality skyrocket the following month). If you work for an OEM, and your dealerships keep begging you for leads every chance they get, who are you not to give them what they want? All sorts of third-party leads can be bought (and sold), plus thousands upon thousands of non-consumer-initiated “leads” generated from phone, website, kiosk, and car show interactions. More is better, right? Then there is the dealership, which has to pay for all these leads in some way, shape, or form and also pay somebody to convert these leads into a transaction, with no guarantee of that customer ever coming back to spend money ever again. Value is truly in the eye of the beholder under these circumstances.
Has Anything Changed Over the Last 10 Years?
Maybe those taking the scientific approach are wrong. With all of the continuous changes in technology and consumer behavior, it’s possible that findings no longer stand up to the test. I already had my theories, but I wanted to prove myself wrong. With that in mind, I took a sample of our dealerships, focusing on the stores with the most consistent performance to minimize the variable of the lead handler on the results. The sampling and analysis methodology is detailed below so readers can interpret their results under similar conditions. Just over 111,000 ADF/XML leads were part of the initial sample, with roughly 1,000 filtered out, making the final sample a touch over 110,000 leads. Of that, 44% were organic, and 56% were non-organic. Here’s what I found:
Quote Payment CTAs
Quote/Payment CTAs generate the most leads, and it’s not even close:
- One out of four organic leads fell into this bucket.
- CTAs of this type generated nearly as many quality leads as the last 14 categories combined.
- These CTAs performed moderately well in other metrics, but were outperformed by more targeted CTAs at the cost of volume.
Availability and Test Drive CTAs
Targeted CTAs regarding availability and test drives generated far fewer leads, but performed exceptionally well across other metrics.
- Test Drive CTAs: While generating just over 1% of all leads, test drive leads generated a better quality lead.
- 70% higher closing ratio than Quote/Payment leads.
- Nearly 40% better set rate.
- 30% better show rate.
- An average-days-to-sale that was 3 times faster.
- Availability CTAs: Availability CTAs generated close to half of the good leads that Quote/Payment leads generated, but resulted in:
- 3 times fewer bad leads.
- Over 45% better closing ratio.
- Handily beat Quote/Payment CTAs in contact percentage, appointment set percentage, and show percentage.
Vague and Overly Specific CTAs
Vague and overly specific CTAs performed underwhelmingly relative to CTAs that indicate a clear intent.
- Contact Us/General: While making up 4.5% of the total lead volume, Contact Us/General CTAs performed poorly across the board.
- Nearly two out of every five leads were rejected as a bad lead (strictly meaning incorrect or insufficient contact information, confirmation of no intent to purchase/accidental submission, or the lead submitter was not legally able to purchase).
- Compared to Quote/Payment CTAs, these leads sold 10 times fewer vehicles, with a 26% worse closing ratio.
- More Information: While this particular CTA didn’t perform that poorly relative to other CTAs, the volume of leads does not suggest it made any difference.
- This CTA drove less than 1% of the total organic leads, of which 84% were marked duplicate.
- The person pushing the button best demonstrates customers will find a way to contact the dealership no matter what.
- Finder/Order: The thought of allowing a customer to request a specific vehicle may make sense, but it doesn’t account for that unit being found or the ability to order that vehicle, for that matter.
- When given the option to click this CTA, interested parties generated less than 1% of the leads (0.27%), which translates to almost 93 times fewer leads than Quote/Payment leads.
- What’s more, nearly 1 out of 2 was marked bad (in the same manner as described above), the closing ratio was worse than Quote/Payment leads to the tune of threefold, with nearly every other performance metric sitting in the worst or second to worst position.
- Super Specific:
- The category marked as Other comprised mostly of leads generated from search bars on the dealership’s homepage or search results page.
- Just over one out of every four of these leads were marked bad, with an overwhelming number of them indicating no intent to purchase whatsoever.
- Of what was left, nearly two of every three could easily be categorized as one of the standard CTAs, meaning the customer could arrange a test drive or ask for a quote either way.
- While those good leads did show promising results across other metrics, Quote/Payment CTA leads generated 20 times more sales with a lot less work.
Other Noteworthy CTA Performance
- Share, certified program-specific, detailed vehicle landing pages, and college-graduate-related CTAs all generated so few leads that they would be considered an insignificant sample size by academic standards.
- Directions, parts, service, and employment CTAs generated more “leads” than some of the retail-centric buttons, but accounted for a whopping 0.10% of the total vehicles sold.
- Pop Up/In/Under CTAs performed significantly worse than most of the CTAs, with more than a four times worse closing ratio compared to Quote/Payment CTAs (0.92% better than the natural closing ratio of a lead if a dealership did nothing), a contact rate that’s nearly half of Quote/Payment CTAs, with nearly every performance metric at the bottom or 2nd to the bottom.
An Objective Opinion
Knowing that some will want to bring AI into the equation, I decided to be proactive to discover what I might be missing. I used Auto Agentic’s data analysis agent to review all the data I used for my spreadsheet, without including any of my calculations. I simply explained to the agent that, “I’m trying to understand what calls to action perform better on dealership websites. I’ve included data below that includes the CTAs/sources, the CTA type, and different KPIs across the first row. Can you please tell me how each CTA type ranks based on the KPIs (Total Leads, Good Leads, Bad Other Leads, Sold in Timeframe %, Sold from Leads %, Avg Days to Sale, Internet Actual Contact %, Internet Avg Attempts to Contact, Appts Set %, Appts Shown %, Avg Days to Appt Set, Avg Days to Initial Visit)? Please use all of your capabilities through the analysis, and please use academic standards.” In all, eight pages of insights were generated, including various implementation strategies. Clearly, this post is already way too long, so I’ll share some of the more interesting points that were raised:
Argument for Fewer CTAs:
- Focus on High-Performing CTAs: By reducing the number of CTAs and focusing on those with the highest performance (e.g., Test Drive, Availability, and Chat), dealerships could potentially increase the overall quality of leads.
- Streamlined Process: Fewer CTAs mean fewer distinct processes for staff to manage, potentially leading to more efficient handling of leads and better resource allocation.
- Reduced Complexity for Customers: A simplified set of options might make it easier for customers to engage, reducing decision paralysis.
- Improved Response Time: With fewer lead sources to monitor, staff could potentially respond more quickly to incoming leads, addressing the bottleneck in lead response time.
- Better Training Focus: Staff training could be more focused and in-depth for a smaller set of lead types, potentially improving conversion rates.
Argument for More CTAs:
- Catering to Diverse Customer Preferences: Different customers prefer different modes of interaction. More CTAs provide options that cater to a wider range of customer preferences.
- Increased Data Collection: More CTAs allow for more granular data collection, which can inform future marketing and sales strategies.
- Opportunity for Micro-Conversions: Additional CTAs might capture leads at different stages of the buying journey, allowing for nurturing of leads that aren’t ready for high-commitment actions like test drives.
- A/B Testing Opportunities: More CTAs provide more opportunities for testing and optimization, potentially leading to discoveries of highly effective new lead sources.
- Competitive Advantage: Offering more ways to connect might provide an edge over competitors with more limited options.
Balanced Approach Recommendation:
Given the observed bottlenecks and resource constraints, I would recommend a balanced approach that leans towards fewer, but more strategically placed CTAs. Here’s why:
- Quality Over Quantity: The data shows that certain CTAs (like Test Drive and Availability) consistently outperform others in terms of conversion rates. Focusing on these high-performing CTAs could lead to better overall results without increasing staff workload.
- Addressing Bottlenecks: One of the main bottlenecks appears to be in the follow-up process, particularly for certain CTA types (e.g., Price Watch has a long time to appointment set). By reducing the number of underperforming CTAs, staff can focus on quickly responding to and nurturing the most promising leads.
- Resource Optimization: With limited staff resources, it’s crucial to make the most of available manpower. Fewer, more effective CTAs allow for better resource allocation and potentially improved response times.
- Improved User Experience: A more focused set of CTAs can lead to a cleaner, less overwhelming user interface on the website, potentially improving overall user experience and conversion rates.
- Targeted Automation: With fewer CTAs to manage, dealerships can invest in more sophisticated automation for initial lead responses and follow-ups, ensuring that no leads fall through the cracks while waiting for human intervention.
As someone who has been analyzing large data sets and demography for nearly 30 years, worked to build an OEM-adopted lead aggregator and scoring system from scratch (personally overseeing third-party lead acquisition for two of the five largest dealer groups in the US as a result), as well as training and guiding thousands of dealership employees across North America for the last 15 years, I was quite impressed that a system that’s a year old could come to my own conclusions in exactly five minutes: People are the limiting factor.
The People Problem
Two very simple and objective truths support this assertion. The first truth is that a human cannot respond to two leads simultaneously. The second truth is that time spent on responding to one lead comes at the potential expense of not responding to another lead. In turn, the responder is always at risk of taking too much time on one lead when another lead could be ready to purchase immediately. This behavior is demonstrated daily at airport coffee shops around the world. The person near the end of the line knows exactly what they want to get, yet the person trying to order doesn’t know what size a “Venti” is or has no clue how ounces convert to milliliters. If the employee at the coffee shop decides not to politely ask the person to step aside to help additional customers, they run the risk of losing those in line who complete their order in just a couple of seconds. It happens all the time.
As many have heard me say from training sessions, an interview, or a stage, every button on your website makes a promise. If you cannot, or will not, fulfill that promise, fewer clickers will become customers. For every different call to action, there needs to be processes to support it. For those processes to work, there must be training to support the processes. For the training to work, there needs to be capable leaders to teach and reinforce the proper techniques. For any of it to work, there needs to be individuals capable of understanding the process, who will embrace the training, and accept the guidance of the leadership. All of this assumes that the management in the value chain will play their part. If any team member throughout the stages of the sale, from initial response to delivery, fails to fulfill the promise, the success of the CTA will diminish.
What Happens if You Fib?
Are you curious why there is a drop-off in conversion regarding Quote/Payment CTAs? In all honesty, we can’t account for all the variables (affordability, creditworthiness, insanity, etc.) that happen on the customer’s side. But if a dealership wanted to account for all of the variables on the showroom’s side, they could measure if the potential customer was sent an email response confirming a price on the vehicle they requested. With that in mind, I used our TaskTeacher platform to analyze how often customers received a quote when it was part of the required process. I found that during the same period as the rest of the data, the customer received some sort of quote or price confirmation only 82% of the time over an average of 9 days. 9 days. Mind you, this is for lead handlers who know they will be graded for pricing activities. This decision was up to the person handling the lead, or their management, to share or not share a price when it was promised on the website. On the other hand, confirming availability and an invitation for a test drive are literally built into the templates, whether it’s on the CTA or not. Is this a coincidence? I think we know the answer.
Decision Fatigue is a Two-Way Street
The fact of the matter is that the wider the variety of CTAs on the website, the heavier the load is placed on each person who is required to respond to those leads. If that person does not know how to respond, or is required to escalate the matter to additional personnel before they can respond, it delays the response to not only that lead, but every subsequent lead that person has. Remember, each person has 56%(!) more leads to respond to that aren’t organic, half of which are OEM leads that can’t be turned off. Suppose that person is also held accountable for the length of their initial response (which is nearly universal). In that case, the person handling the response is forced to decide to send something that is less than optimal if it goes beyond something they can easily handle on their own. If the CTA takes more than what the responder is empowered to do autonomously or is beyond their training, the value of a CTA is diminished relative to its impact on higher-performing CTAs. More decisions, more problems.
Keep it Simple for Everyone
Not everything has to be a giant button. Amazon, along with other retailers, does include text links throughout the body of the page to call out certain features, options, or opportunities that may have more niche appeal. If a dealership wanted to more specifically measure certain behaviors, create a more targeted follow-up, or merely satisfy OEM requirements, the text links would provide an additional option for the customer who is taking their time. Specific pages could also have unique CTAs relevant to the content, allowing shoppers an easier choice to get into the right conversion funnel.
If leads for the sake of leads isn’t your primary goal, and your dealership doesn’t have endless resources to hire, train, and maintain the optimum amount of staff, then it’s probably best to stick to the absolute minimum amount of the CTAs your OEM requires for both the SRP and VDPs. Although you may not have the best staff, training, and management to handle those opportunities, it’s much easier to build a small handful of consistent processes, as opposed to more than a dozen processes, that require even more training, measurement, and potential managerial requirements. Nothing could be more costly than trading a customer willing to buy today for someone who has no intention of returning your contact, or worse yet, having to tell someone that you can’t keep the promise your button made.
Social media can be fertile ground for good discussion. However, the limited nature of the length of response can often overshadow the actual intent of what’s being written. However, when scientific methods are called into question, then someone has to take the responsibility to explain their methodology so that someone else can compare their results in the same manner. Although no two dealerships are truly created equal, we can take steps to share our findings in a way that’s repeatable for others to test. If you find that having more CTAs is worth it to your organization, then that’s up to you to decide. However, if you’d like to test to see if a streamlined pathway to conversion works better for you, then I encourage you to run similar experiments and share your results.
Whether it’s part of a comment or written on a button, remember, words matter.
*Research Methodology
The dealerships consist of metropolitan, exurban, and rural dealerships with a balance of foreign and domestic brands, excluding niche, high-line, or exotic brands. The dealerships are not in close proximity to manufacturers or suppliers. A full 12 months of data (1/24-1/25) were used to neutralize the effects of seasonal fluctuations, incentives, and inventory disruptions. Data was collected through aggregate/enterprise CRM reporting with standard availability through the CRM provider. Sources were filtered using a methodology that included only leads originating from ADF/XML. The sample also excluded any sources that indicated performance results without indicating a valid lead or indicated performance results that exceeded 100%, which is indicative of improper data collection. Since there are no standards on how a lead is interpreted by the CRM, CTAs were categorized based on the most logical category based on source name and/or vendor name. All precautions were taken to ensure that data was analyzed in the most objective manner.
