This isn’t your average success story. It’s not a neatly polished case study that only gives high percentages, lots of ROI and a super happy customer. No, this story is about a struggle, a learning process and how we became better at our jobs.
Scene 1: The Client Talks to 434
Grizzlyhive Products* approached 434 Marketing with a need: Their eCommerce website was roughly 20 years old and, in that 20 year span, they hadn’t seen any increase in online revenue. Knowing that 434 Marketing is the master in all things digital marketing and web design (and other things, too), Grizzlyhive Products gave us a call.
Before giving them a quote on the development of a new eCommerce store, we did our due diligence: We researched their competitors, we evaluated the eCommerce user experience, we studied Google Analytics and Webmaster Tools data. When it came time to present the client with our recommendations, we had a very clear idea of how the website should be laid out and how the overall experience should flow.
We kicked the project off in May of 2017, with four primary goals in mind:
- Increase website traffic.
- Improving the rate of conversion (Traffic to Revenue).
- Give the website a fresh, more current look.
- Organize the huge selection of products to make it more successful and user-friendly.
*Real client name redacted to protect their privacy (also, it’s no one’s business).
Scene 2: The Next Steps
Over the years, we have built a seamless process for website design and development at 434 Marketing. After our kick-off meeting with the client, we immediately went to work on more design research, so we would understand exactly what the aesthetics of the website should look like to draw in the client’s target customer demographic.
By the middle of June, we presented our design recommendations to the client. Approval took no time at all, and we entered the throes of development.
We had an incredibly clear path that we wanted the user to follow and it was neatly laid out by our navigation structure, calls to action and overall buying experience. The client was thrilled! Development, while a lengthy process due to the 800+ products listed on the site, was smooth.
In early September 2017, the site went live. We kept the old site live, but not visible, just as a fail-safe backup in case we needed any old product information or previous revenue statistics. However, we set up redirects so the user would never see the old site again unless they had the exact URL.
Scene 3: The Battle for Revenue and Google’s Cold Shoulder
When you launch a new website, you should initially expect to see a 5% to 10% decrease in organic traffic. That’s because Google re-evaluates this new sitemap, content and structure and updates its trust and score of your website. It does this by crawling and analyzing the content, user engagement, page speed and dozens, if not hundreds, of other factors following the website launch.
It’s entirely normal to see your average position for high value queries to drop but, after a few weeks, that average typically levels back out to normal. The only time that that should vary is if you change the root domain along with your overall website design and structure. This is because your new root domain has no history whatsoever in search engines and claims a much lower domain authority, which is high on Google’s rank list.
But…levels for Grizzlyhive didn’t normalize after a few weeks. In the month following launch, we saw a spike in pageviews, pages per session, average session duration and a significant improvement in bounce rate (year over year comparisons).
But, as that month continued to progress and we moved into November and December (historically this eCommerce store’s busiest time of the year), we kept becoming more and more aware that something was off. Way off.
Average users were down by roughly 24%—to the tune of 2,600 customers and overall sessions by roughly 13% (or 1,700 sessions per month). As we dove deeper into that analytical data, we were seeing that all traffic channels (direct, referral, email, social and paid) were up by decent margins. However, organic traffic was down by 34% (November 2016 compared to November 2017).
What that equated to was a 16.67% drop in revenue in November and a 34.73% drop in revenue in December (year over year comparisons).
We’d like to say that we were the only ones feeling the pain of those decreases but, if you’re a business owner or have spent any time at all around business owners, you will know without a doubt that it wasn’t the case.
We had countless conversations with our client over those few months. Both parties weren’t sleeping at this point and we were left wondering what the heck we did wrong?
We implemented additional marketing tactics at no cost to the client to get him over the holiday shopping season relatively unscathed, but we both knew that that wasn’t a long-term or healthy marketing solution.
For some reason, Google was giving our client’s new website the cold shoulder.
We were tracking specific query data with keywords and phrases that we knew were driving users to the old site, and those impressions were even down.
Scene 4: 434 Makes a Plan
We fiercely fight for our clients. We understood the pain, frustration and confusion of losing 20%+ in revenue month-over-month and, while the client wasn’t even asking for us to “fix it”, we knew we had to.
So, we made a plan.
We matched the queries from before the site shipped with the queries from after the site shipped. Out of 1,000 queries, 262 matched—which meant that similar keywords and phrases were still driving traffic to the website. The queries that did match? They accounted for 77% of clicks to the site. That was the good news.
The bad news? Impressions, clicks and click-through-rate were down…down…down. (A snippet below, keywords and phrases have been removed):
We took that information and evaluated the areas where we could see the fastest level of improvement. We made a few recommendations:
- Increase the word count on product descriptions to at least 500 words
- Remove blog content with little to no traffic
- Level up top performing blog posts
- Improve site speed issues
- Create content-rich blogs of 950 words or more
All parties agreed to that direction and we all had our marching orders.
Over the next month, both the 434 team and the client wrote like our lives depended on it (and, for some of us, it did). We increased product description counts, wrote lengthy but valuable blog content and corrected site speed issues on 59 pages.
Then, we ran some more tests to see how the site was performing.
Scene 5: Things look Bleak
After evaluating our changes and the effects they had on the client’s overall positioning and traffic, we saw some improvements, but not enough to write home about. Sure, the site was beginning to rank for the keywords that we were working to focus on, but it was still nowhere near where the old site was.
But, then, Google (as Google does) swooped in with a Search Console update that changed the game.
Offering us 16 months of data, we were able to look even deeper into the website’s overall performance compared to the previous website. Armed with a considerable amount more data than just a week before, we dissected both on-the-page factors and off-the-page factors that could be helping or harming the website’s search success.
And here’s what we noticed: There was a section of content that was being crawled on the old website that we weren’t aware of.
You may find yourself at this point recalling a bit of foreshadowing from our story so far. Here’s a refresher:
In early September 2017, the site went live. We kept the old site live, but not visible, just as a safe-fail backup in case we needed any old product information or previous revenue statistics. However, we set up redirects so the user would never see the old site again unless they had the exact URL and assumed the previous developer updated the robots.txt file with a no-index command.
That’s right. Our assumption that the robots.txt file had been updated on the old site by the previous developer, was wrong. The website was still being indexed and cached by Google. This tiny adjustment (literally a button click and standard protocol for a staging site at 434), meant that Google was still crawling the old site and offering it as a result within the Search Engine Results Page (SERPs). However, we were wrong.
In fact, Google regularly downloads robots.txt content. It won’t download it every time it wants to look at a page, but it will routinely cache the content. So, we got back in touch with the website’s previous developer and requested that the old site be archived entirely.
In just a few weeks time, we saw a 12% increase in organic traffic and a 70% increase in revenue from organic traffic.
End Scene: 434 Marketing Learns…A Lot
So, what did we actually learn from all of this? A lot. But to summarize:
- When you ship a new site, even if you make the necessary tags to not index a site, Google may still index the site. Archive the old one as soon as you can or ensure your robots.txt includes no index commands.
- Site speed matters more than you can even imagine. Shoot for 3 seconds of load time.
- You can’t put enough value on content. Write well, write a lot and write to what you know your users are searching for.
- We have some of the most trusting, patient customers in the business. We are humbled and grateful to them every single day for taking marketing journeys with us and allowing us to correct mistakes we don’t even know we make. But, we will make it right…we will always make it right.
If you have an eCommerce site that needs refreshing (or even just a normal, static website), don’t let this sob story stand in your way. At 434 Marketing, we are constantly adapting to new trends, algorithms and regulations set forth by the internet powers and will work tirelessly to make sure your marketing is a success. Get in touch. Let’s chat.