We’ll be looking into some real-life examples of projects I’ve worked on, some solo and others with my team. Some were a sweet stroll to growth and others were downright abysmal. I hope that by sharing some of my experiences as an SEO at the helm, it will help you to recognize potential issues so as to avoid them in the future, whether client-side or agency.

Real names will not be used and some details have been changed to protect the innocent. Many of these cases happened while leading my own agency with clients in San Francisco and New York. I was the primary SEO for all of the accounts.

Each case study ends with what I like to call lessons learned.

Multimedia Entertainment Site

The first case study is a tale of a multimedia website that offered videos featuring up-and-coming artists from many disciplines. There was little to no text at all, which presented some unique challenges. Revenue for the entertainment site came from ad monetization. SEO was their ultimate desire, because it represented an especially lucrative flow of cash for the company; to them, it was a pure gravy train with biscuit wheels. Also, ad traffic is fairly hard to resell (or upsell) because there’s often a lag in data collection.

When we took over the website’s management, the first step was to set up analytics instrumentation properly, then monitor traffic coming in and out. The website generated revenue for artists and advertisers mainly by the sheer amount of traffic the site got. The team who built it came from Hollywood and they had just the “right” network to get buzz. However, advertisers on the website complained that they were getting little to no value from the traffic “visiting” their ads. They were experiencing “record-low conversions,” as in close-to-zero conversions. Even on a bad day, there should be a few conversions to a site that brings traffic—1% at a minimum.

Typically, sites that resell their traffic want to do so with organic, because the ad revenue is pure gravy in that scenario. You don’t have to pay for organic traffic long term if you can maintain rankings, so you’re essentially reselling your real estate. I liken it to owning a home and collecting rent from tenants. Also, traffic that originates from organic will tend to stay on the site for a much longer time. The upshot is that this kind of increased engagement is almost impossible to buy.

The culprit in this particular case was that the site’s owners were purchasing ultra cheap paid traffic in order to resell it at a higher rate. Upon further examination, I was able to quickly determine that all of the traffic coming to the site was being sent from subdomains. I tried to visit those URLs, but they weren’t anywhere close to real. Most web referral traffic comes from sites like reddit.com or t.co. The URLs can vary, but it’s somewhat rare to see a subdomain providing traffic from the outset unless it’s part of a campaign.

The questionable traffic was not only originating from untrustworthy URLs, the average “user” only spent three seconds or less “using” the site. Even with mobile traffic, there’s just no way that any human will click off of a site within one second, much less thousands of them in a short amount of time. This didn’t feel like a very likely usage pattern. So often with technology, the issues boil down to the same adage: Garbage In, Garbage Out. It became very clear to us that someone in the chain of custody was purchasing shady bot traffic (unknowingly, we hoped) and then reselling the traffic to advertisers. Sometimes the stories didn’t quite add up when we asked about the details of the traffic purchases.

A WARNING ABOUT BOUNCE RATES

Session duration is a deceiving statistic in the case of a bounce. The user could have spent 30 seconds or 30 minutes on the page, there’s no way to know definitively.

Analytics suites like Google Analytics can only measure session duration up until the last event that it captures. If a user lands on a web page, and then clicks the back button, the tool has no way to know how long the user was on the page. This can lead to unusually small session duration statistics when, in reality, the problem is in the capture of the data. One way to mitigate this is to record additional events such as how far a user has scrolled. By recording events beyond just page views, Google Analytics can have a more accurate basis for its measurement of session duration.

Let’s just say that all the emails flying around with explanations of the fake traffic were equally as confounding. We were never 100% sure at what level the deception occurred (i.e., if the client knew the traffic was not coming from legitimate sources). If you’re using a broker to buy third-party ads, it is best to make sure they operate above board.

Lesson Learned: If traffic coming from a website looks as though it is not real, then simply visit the site’s listed URL. If you are receiving traffic from a website that doesn’t exist in a browser, it could be a redirect or something else nonlegitimate. There are many companies trying to broker traffic deals. In many cases, the traffic is not human; it’s essentially scripts or bot hackers that are run to inflate numbers, and therefore brings zero value to the site. In fact, too many bots can be a sign of trouble for an organization’s overall health.

Low Confidence, High Traffic

This next case study is for a (once) celebrated enterprise middleware services company. Their revenue model was to offer a free limited trial period. They captured users and converted them to monthly subscriptions thereafter, which were tiered. I was brought in but quickly informed that life was good because leads were “rolling in hot.” There were some kinda-sorta-maybe looming concerns. First of all, the organization was unaware of what sort of ROI they were getting for their marketing dollars.

All kinds of information about their referral traffic was flat-out missing. The information pipeline was not merely inefficient; it was actually unknown in some areas. Therefore, they were crippled in both decision-making and planning. Because their data was too murky, they found it difficult to trust their own conclusions. Rather than looking through a murky glass, they needed some transparency. After all, data-driven decisions are what every modern organization strives for.

When the project began, cookies were fender-bending everywhere. Some of the Google Analytics UA property numbers were actually repeated three to five times in the same page. Why? Even the structure of the site became unbelievably complicated with the way that password walls were put up on the site. Complications became even more baroque when examining the setup of account and password creation. It was hard to understand where the gates were because there was no logic to it. You’d click on random things and be asked to sign in, sometimes to different systems! It was unlike any mess I’d ever seen in terms of data and general user experience.

There were over 100 full access administrators in the master analytics account. Many of the accounts (again, listed as admins) turned out to be duplicates. There were duplicates because people wanted their personal Gmail addresses in there, too. We matched over 25 names of people having both their work and personal mail as admins. They were unwilling to open another browser with their “work email” to view Google Analytics at home. This type of cavalier attitude towards building process and security led to sloppy and chaotic data collection.

When I asked seemingly simple questions in initial meetings, there were many excuses, but mostly visible annoyance, some furrowed brows, and lots of confusion. Attitudes eventually became so casual in the marketing department that almost nobody knew where traffic, users, or revenue were coming from.

In the first meeting, it became clear that we were 100% on our own to construct the picture of how this Franken-site came to be and how to fix the data leaks. Physically viewing the source of a website is generally where you want to start when looking to validate data capture. Thoroughly examine and read the site source like a novel for all top-level web pages and do not ignore the outliers.

An outlier in this case would mean custom scripts and hacks to the tracking scripts. If a routine header JavaScript filename is customized with some cutesy name, for example, you can assume someone meant for it to be there (i.e., customization has happened). Customization is a killer in a web development organization, because it facilitates mistakes and creates support dependencies. Another more subtle and hard-to-catch issue could be that, as the SEO, you run a speed test and find that something seems to hang on load. Upon further examination, you see JavaScript files with custom-feeling names. Uh-oh, time to investigate.

Every department was able to deploy nearly anything they wanted for their individual product lines. Multiple teams of web developers were updating the site without one central repository, which was the biggest reason for the physical audit of source code. Nobody seemed to notice how each new product’s destination page felt like it was a new site. To click through each top-level page was a usability nightmare from the late 1990s; the headers didn’t match up and sometimes the colors changed. Every page hopped to a different URL path, and some URLs jumped to subdomains. What I mean is that each new product or feature got a subdomain; sometimes an independent CMS was stitched up to it, sometimes not. When you engage with too many subdomains and CDNs, there’ll be a headache. When teams of developers build individually customized sites like some kind of Winchester Mystery House, you are sure to have epic junk in the trunk.

Too many handshakes were happening all around. An NGINX was hooking up to one side of the site, whereas Elastic Load Balances (ELBs) were hitched to the other. The server hand-offs were seemingly random and therefore the data was too. Let’s just say it was an anarchist’s data rodeo given the countless servers and services involved. Once I recognized the massive overlapping in tracking scripts on individual pages, we were able to account for the discrepancies and begin to map it accordingly.

I created a spreadsheet with each top-level page URL on the site and its subpages (because they were all different). Slowly we were mapping which UA property numbers were being used and in which version of the tracking script. Very few properties were using universal analytics or tags at that point in time, so developers really liked to add their own tweaks to the scripts. Some of those tweaks included fun things like taking the tracking script and placing it in an external file used in random places or adding multiple UA numbers to the same script.

Cookie collisions are real! Sometimes the desire is there for developers to tweak. Unfortunately, the number of leads showing up as newly created accounts in GA reported more than double what was actually happening. Since nobody was actually following up on the leads, the fact that the machine broke didn’t even really matter to anyone.

We helped them get ready for the acquisition by scrapping nearly all of the separate properties and bringing the data rodeo to an end. Too much complication with subdomains meant more upkeep for them. There’s nothing wrong with using subdomains, but there should be a solid reason to use them (other than a desire to not engage in communication person-to-person).

On a macro level, the ELBs were improperly configured. They had so many handshakes for different country sites that issues became difficult to replicate. Outages were close to impossible to fix because nobody knew where to start. A rat’s nest of server topology was holding the organization back until we were able to migrate to a more stable configuration.

Everything ended up working out well for the acquisition, and data was finally able to stand up to scrutiny. With the ability to track marketing activity properly via campaigns in analytics, the company could properly determine what marketing activities brought revenue.

Lesson learned: Too many handshakes lead to headaches. Just because you can customize a tracking script doesn’t mean you should. Document and communicate about major site changes. Do not go nuts on deploying subdomains. Complexity always grows faster than you anticipate. Avoid it if you can. Eventually, if you rely on other tools too much, you’re just giving away your content while confusing users. Over-customization rarely scales well in a web development environment. Follow leading metrics, not trailing ones.

Toxicity in Search City

Some industries are so big that the top 5% is cutthroat warfare because millions are at stake. I’ve only seen a handful of industries where the competing companies routinely hit the CPC maximum of $1,000 per click. If a company is operating at the level where a qualified visit can cost roughly $1,500 or more, it’s safe to say that anything can happen.

When competing for highly coveted SEO terms, it’s important to diversify and also focus on organic search. The company in question in this case study did not hesitate to deploy numerous SEO experts to try to dominate organic rankings. Turns out the competition did, too.

Once we were engaged as their research SEO firm, we quickly noticed that many new industry sites were popping up instantaneously. What we found even more interesting was that the new sites were linking to our client. As if by magic, overnight, we were the most popular girl at the prom. Unfortunately, the sites were overwhelmingly spammy in nature, as many contained the same duplicate content and weird spelling errors. We did not luck out. Someone was sinking our boat! Too many shady traffic referrers can sink a website’s rankings when left unchecked.

We used the whois search from domain tools to determine the main locations and hosts used by the spammy pages. The hosts were overseas and many of the registrar listings were anonymized. The nonanonymized registrar listings were to a dummy company with no phone number or physical address anywhere online. It was tempting to go and find a “hacker” of some type, but unfortunately that’s not how the real world works. Even if we could have done some TV-like hacker magic and discovered information about the offending company, we would have put the client at risk by breaking the rules. Finding these people was a dead end for us, best pursued by the legal team.

As soon as we spotted the spam coming (obfuscation), the keyword rankings started to slip. Oh noes! Given the algorithm changes enacted earlier that year, we were correct that the client’s volume of traffic would degrade quickly once these sites started to multiply. Each day, more sites materialized frenetically, all containing duplicate content along with other spammy text and links.

Fortunately, we discovered the issue before any real damage had been done. We were able to disavow the toxic links quickly and thoroughly using Google Webmaster Tools. It took several rounds of submitting the disavowals, a strongly worded letter to the competitor, and another nicely worded letter to the search engines to wipe the links away, but eventually the databases were updated. When all was said and done, it took a few months.

Lesson Learned: Do not forget to look at sites who are linking to you. When stakes are high enough and competition is steep, anything can happen—even warfare on your organic rankings. You’d be surprised by the lows that normally ethical companies will sink to when prosperity hangs in the balance. Take the high road.

Special Snowflakes

Once we worked with a client that had the opportunity to create a new name for their flagship product after an acquisition. The only stipulation they had was that they had to change their company name completely. They were allowed to use virtually any name they wanted. This meant we were working with a new website that stood alone. We needed this page to be able to stand and make sense on it’s own, but also appear as part of a corporation (with their other brands). There were almost too many directives for us to parse, and lots of confusion at the top. So we were told by the company that the site simultaneously needed to stand apart, but be recognized as part of a bigger brand. An attitude of entitlement doesn’t achieve winning results; it drains resources. This project was an interesting case of woeful ignorance, with many folks on the marketing team out of touch. Ambition is generally a positive force for a program, but not when the champagne wishes exceed SEO dreams.

The client chose a single word for their new name, which was also being used by three other companies. When we were discussing domain names, I warned that the SERPs themselves were confused about what to serve up for this particular branded query. The SERPs showed all three of the aforementioned companies first. The name was also a common word used in normal day-to-day speech.

As an SEO, it’s hard to take on a project like this one because winning the war is not likely to happen. We warned continuously that to own branded search on page one would be a multiyear long-term project.

Despite my best effort to be realistic, the company still felt that they should rank on page one the same day the new site launched. We tried on multiple occasions to set expectations properly, but they felt they should just “own it.” As a temporary means to reinforce the brand efforts, we used pay-per-click (PPC) for a short time. Even with a highly locked-down keyword restriction, the budget was expended in an hour or less. The quality score on the ad was fairly low since most of the people using that term had different aims. We saw painfully low conversions on branded search.

The PPC budget was a little more than $20,000 per month, which wasn’t making a dent nationally at that time. So we pivoted. The short-term strategy we devised included leveraging paid social and old-fashioned elbow grease to get them traffic with social bookmarking, YouTube, Reddit, email, etc. Another thing we had working in our favor was that the client had a funny bone and let us take their social media out for a walk. Eventually the social links for Facebook pages and YouTube numbers started to look like what we wanted them to. We pulled out all the stops and grew a serious fanbase on social media, which was important for their industry. The company finally owned their name!

Lesson Learned: Sometimes the best battle strategy is to not to fight on the battlefield you’re told to. There are rare cases when it does not make fiscal sense to build an SEO program for a business. Getting traffic funneled through social media should be considered a win, but goals don’t always make sense. You have the power to change and level the search battlefield. It is more than OK to leverage paid tools like ad retargeting if they make more sense for a project.

Too Cool for School

This project was for a privately run site that offered online training for fun and profit. Their curriculum provided students with professional hours toward certifications in many different disciplines. Somehow along the way they’d created some extremely vocal detractors online. They came to us when they had another SEO firm still engaged, because they didn’t trust their firm yet. There was something about the kick-off that didn’t seem quite right. They immediately sent us all the working documents that the other firm had created. This didn’t feel like a good thing to us—we didn’t really want or need to see a competitor’s work product. But alas, unraveling the mystery began.

The content was highly technical; the SEO program was too, run 100% by spreadsheet. A previous agency created a mobile app that had very few features and yet still didn’t work very well. Pride prevented them from end-of-lifing their failed native app, but that’s not what became the problem. Someone who was once closely related to the company (at the top) had serious drama.

The issues we faced in our search battle were mostly bad reviews and a hint of social rage. We were assured that the company had released this person from their employ. Our research online gave us no reason to doubt the client’s story that they’d simply hired the wrong person. What we had in front of us was not a salacious or sexy scandal, and uncovering the trail of information was easy. After some research we uncovered umpteen people across the country who were burned badly by this person, who we’ll call Trolly McTrollTroll.

One person in particular just couldn’t let it go—they wanted to burn the internet down. I would call them the epicenter troll living under our reputational bridge. Page one for a branded search of this client’s name returned massively bad reviews, an active Twitter account with a similar name, a Quora post, and some blogs. Some members of the executive team took the online commentary personally. One of them decided to respond in hopes of helping the school, but it had the opposite effect.

The client had a surprisingly small and inactive social presence for an online school so we worked to build the social accounts up quickly. Building up the accounts meant simple stuff at first like filling out fields, posting interesting updates, liking other people, etc. We registered for listings on sites they didn’t think of before, like individual professional associations for the subject areas. Flooding out the bad with good is often the best strategy when negative information you’re trying to suppress is actually incorrect or unfounded.

Instead of continuing down the same old PPC branded campaign road, we heavily researched the organic queries. Many reputational queries were creeping in and they had low click-through rates. We added new terms to the PPC campaigns that addressed potential students who sought to understand more about the school’s reputation. Next, for a customized landing page, we crafted text with targeted copy to build up reputation: quotes from graduates, press mentions with logos, and some text in the warmest font we could find to reinforce the school’s position. Warm and smooth fonts to use on landing pages are Georgia, Minion Pro, or Garamond.

The school maintained an extensive list of alumni email addresses, so we ran campaigns. Some of the campaigns started out as simple surveys to gauge sentiment. Once the list was segmented out properly, we asked the students who seemed to be net promoters to write a review if they had a positive experience.

In terms of reputation management we were reporting all Tweets and handles that violated Twitter’s policies, including those we believed to be impersonations or bots. On Facebook, we banned people violating profanity rules or those who we believed to be our troll or subsidiary trolls. Spam reports were filed with various hosts where we could see duplicate or hate-filled comments posted. Bad reviews and commentary will vary by site, but once you’ve read enough fake reviews, you start to see patterns.

When we started to take the above factors into account, we knew it was likely the same person or group of people was leaving the bad reviews. We were asked to deploy “whatever methods were necessary” to search out the online detractor’s true identity. After one millisecond of consideration, I declined. We instead suggested they consult an attorney, because at that point it was out of our hands. We eventually turned some of the sentiment around and attention drifted away from the controversy.

Lesson Learned: Always expend the effort to fully understand the goals of a search program up front and trust your instincts if something seems awry. Acknowledgement of or communication of any variety with a troll validates them (in their mind). Responding to a hateful tweet is tantalizing in the moment but only serves to make situations worse. Focus on building out better content, tuning social, and leveraging attention for positive things that are happening. Heck, knock yourself out and throw a press release out there too—it can sometimes trick news results.

Classic Growth Hackin’

This project was one of those during which we kept high-fiving constantly because the money was rolling in steadily. When the chemistry is right for both agency and client, magical things can happen. Strong communication between marketing and product development will only benefit sales. It behooves the growth hacker to stay as granular as possible with sales processes (to identify opportunities). The growth challenges we faced were related to the consumer’s trust. Previous regimes at the company tried and failed at paid search, paid social, and old-fashioned PR (smiling and dialing).

This project was one of the fastest increases in growth for both program development and revenue that I’ve ever seen. We quadrupled the company’s sales in less than two months by leveraging organic, social, and paid traffic channels. We were promoting a business-to-consumer widget that was a cool modern twist on an old established household product. They’d raised a plethora of money from a crowdfunding site. A slew of high-priced consultants and CMOs had been in and out of the organization, which was apparent by their use of four distinct analytics suites of tools. To reach for the stars is a noble goal, but to build momentum teams have to be unified.

At first the product was hard for people understand. After a few months of concerted effort between agency and client, the consumers couldn’t get enough of the product. The spike in sales rapidly led to a backordered situation, which was both a blessing and a curse. We were told that the investors were not going to extend any lines of credit, so once cash was out, it was out. The stakes grew increasingly steep for this client and fortunately we came through for them with mostly organic traffic when marketing budgets got cut. We worked social media super hard seven days a week and pumped out at least two decent blog posts. The blogs were listicles and we spent the time to have really good-looking pictures. We were liberal at times with the gifs and the internet loved it.

Like many companies, they experienced some delays in manufacturing. A natural disaster caused a major part of the delay, so the response by people on social was kind (at first). Promises were made by the CEO and they felt sincere, because I believe they were. The company still needed to generate income while the delays were happening. Only one person worked in marketing for a very long time and more recently a social media manager had been brought in to alleviate the pressure.

We immediately shut down the spending on unused tools, ineffective ads, and vanity paid-search spending. There were glossy ads with no real call to action or offer tied to them. There were even ads for just the CEO that basically only seemed to indicate how great he was as a leader. Needless to say, the ads did not perform well and therefore failed to bring in revenue. Company data was also being shared quite frivolously; nobody in the marketing department even knew who owned the access to most tools. Eventually we found out that several previous consultants actually owned the logins. We also had to calm down an overzealous social media manager who was deleting Facebook comments and responding directly to trolls on Twitter (using the company handle).

Sometimes when social managers read comments about the company they’re defending long enough, they start to take the attacks personally. In this case, the social media manager felt any negative comments on social meant a risk to his job. It was clear that he was simply spooked by a few of the comments and felt backed into a corner for some reason. We recommended someone more experienced (with the product) from customer support take over social comment monitoring and response. Things went more smoothly from that point on.

The organic program was nonexistent; it was hard to figure out who the company was and who it was about. There was zero information on the founder or the CEO of the company, and not even one picture of the office could be found on the public website. We had to email and ask for the address of the kick-off meeting because it was impossible to find, along with a phone number. I can only imagine how someone would feel if they’d bought a product from this company and had any kind of issue. The company was hiding from the very people it had taken funds from.

We came up with three distinct campaigns that we executed and then measured carefully. The first campaign was to showcase the modern design feats of the product and how it was outperforming its predecessor, the second was a loyalty discount for a new product, and the third was a contest to do silly things and then win free stuff. People love free stuff! The contest campaign outperformed the others 4 to 1 in terms of interest.

Lastly we started a very successful affiliate program that included key social and industry influencers. We were lucky to snare a few big industry names, which tipped the scales in terms of brand power. All endorsements used the #sponsored tag so that it was 100% transparent that both affiliate influencers and customer influencers were paid for the endorsements. We gave them custom tracking URLs so we could measure by campaign what was working and what was not. By associating this product with powerful and influential people, we were able to grow the company’s social followings so quickly that all ads became unnecessary.

Lesson Learned: Help wherever it’s needed, even if it starts with providing the path to cutting unnecessary expenses (if a company doesn’t have the confidence to do that on its own). A slow and steady hand sends the tweets. When leveraging influencers, don’t parade them about; treat them like a partner, which means hearing what they suggest. If they’re being paid for their endorsement, make that clear.

B2B Social Spice

This client had a business-to-business tool that helped manage web services. Sometimes individual consultants would purchase the tool, but very rarely. The previous regime had made a total mess of this client’s marketing program. Creatives were painfully bad stock photos that appealed to nobody and hence were performing poorly. The first step was to clean things up and then we got to take a step out with our new site.

The last person in charge of PPC moved over to HTTPS, rightly so, but did not actually change the URLs in the ads console. Every landing page was served up, kinda, but with a broken stylesheet. The conversion rates were the worst we’d seen. Catching that mistake quickly quite literally saved the organization quickly. They were just in the middle of closing a funding round and sales mysteriously started to tank. Questions were being asked and they needed to be answered quickly. The ousted CMO left a trail of digital dust in her wake.

I always take the time to check the physical URLs a company is putting out there, not what they send me. I click on the URLs in any and all social listings, consoles, website footers, etc. You’d be surprised how often things are broken or simply do not match up. It always pays to review what’s physically being put out there on the web and also within any of the promotional tools used against what is believed to be out there.

When clients are not making money because of someone’s malfeasance, it’s typical to demand a quick result from the next regime walking in the door. For us, the quick win was spotting the bad landing page URLs because fixing that immediately improved conversions. The previous consultant also forgot to separate branded from nonbranded search within the Google paid search groups. It’s lesson number one in paid campaigns to start big and then whittle things down. Always separate ad groups intelligently so that you’re able to drive the spend down for the group. Once paid was conquered, we started reviewing the organic side of the house, which had issues too. Every meta description was the same; artwork was generic and tired. It seemed as though every picture was from a classic hackneyed B2B stock art site: handshakes, the “we are all enjoying ourselves sitting and bantering” shot, or a grid with a finger pointing to it (Figure 9-1). Companies often use this type of corporate stock art to convey a sense of professionalism or appear larger than they are. Quite often these types of stock corporate photos come with templates for WordPress website themes.

Figure 9-1. An example of the classic finger or hands on a grid stock art shots. Credit: Shutterstock.

We needed to change most of the visuals quickly while getting the site in compliance with modern web standards. An old Joomla installation made site updates difficult, so migration to a more modern tool had to happen. Once the artwork was updated to interesting images that were originally designed by a designer, the conversions improved exponentially in the first few days.

The ecommerce plugin solution they were using was wiping lead source data out entirely. All traffic showed up as “direct” and not available for 80% of sales. Both the cart solutions analytics and the website’s analytics showed the same thing; our traffic was sourced as “unavailable/direct.” This was a big problem! We escalated to the third party what was happening and it took several weeks to work things out. The issue was with the third-party shopping cart’s site, but we were responsible for fixing it. Ultimately we had to use Google Analytics’ tag manager to pass on the cookie data in the transfer between cart and site.

While this case has the earmarks of your classic case of SEO—clean it up and watch it grow marketing—that is not where we saw our best sales success.

The majority of the clientele for this client were developers in Silicon Valley. They needed to get inside the hearts and minds of the web development crowd, the people whose lives were less pleasant without this particular tool. It’s very hard to snare a developer with a B2B marketing website with traditional trappings like white paper downloads, ask an engineer, or a free webinar.

Because of our own proximity to this market, we knew that conferences for marketing people were taking the side stage to conferences targeting developers. Giant symposiums were forming around getting in front of the developer, face to face. Many more modern developer or maker conferences have areas for origami, crafting, or working with 3D printers. Developers are given a community to call home and camaraderie at these events, so to interject into them takes finesse. Our approach was geek humor and general supportiveness. If someone complained to the conference about allergies, we would offer them a tissue. On social, we were like a cool buddy from high school that’s always there for you, not a company trying to sell itself.

We began to attend more and more developer-focused conferences. The client trusted us after dramatically turning around sales, so we got to tweet for the client account. As a former web developer, there was a certain amount of kidding around I felt comfortable knowing I could do. We’d built up that trust, which is a beautiful thing. I knew my three-letter acronym (TLA) game was tight, so onward I tweeted. We discovered very, very early in analytics that on the days when we attended conferences the client’s site received more leads. Developer leads! We were copied on all leads coming into the site with coded email addresses for filtering. By having insights into the lead flow, we could tell what was working and annotate quickly.

Our assets went from webinars to pushing more and more conferences in real life (IRL). Sometimes intricate tools are best sold to their niches with a human face and interaction or a little personality on Twitter. By being a good guest at the conference, we often received dozens of retweets and favorites. Ahead of the conference, we’d write a highlights piece with background information for particular panels that were targeted to us. One year at the popular event conference, hackathon, and startup competition TechCrunch Disrupt, we had our most significant traffic spike with thousands of views in just an hour. On the other side, we’d get a recap of the conference panel with embedded Tweets and get that going for the next day. Timeliness of content when tied to social and real-life events was our rocket ship ride to developer snaring.

LinkedIn was our secondary community for lead generation, which surprised us. Our experiences on LinkedIn were somewhat negative on the paid side after the company went public. LinkedIn’s InMails had a painfully low click-through rate. We were told our results were poor because we had the wrong audience or we weren’t paying enough. It doesn’t inspire confidence to know that if you haven’t gotten us money in the past that more money will lead to different results. In this case, organic optimizations, highly specialized content, and social promotion benefited the client far more than paid.

Lesson Learned: Know your audience. If at first you don’t succeed, test and test again. Online success can almost always be boosted by coordinated offline activity. Don’t follow convention; follow instincts.

Dysfunction Junction

Sometimes the institutional goals are just a thin political charade. People often bring trouble from their personal lives into work; in this case, it ran rampant. This organization had a lot of things going for it internally, a decent product offering in a space that was exploding. This company was a B2B services industry platform that combined their software with consulting services. The end result for this company’s customers was dramatic savings and increased sales efficiency. To onboard this company for a large enterprise was a no-brainer; they integrated with a good amount of services that were already out there.

We began this project with the excitement all new projects deserve. After studying everything, we had some very practical priority suggestions to get the ball rolling, like fixing broken links and improving the analytics instrumentation. All title attributes were the same for every single page of the website. They were passing traffic back and forth to subdomains for no real reason. It was a bit of a nest. But once we ran out of the recommendation zone, the trouble started.

There was no system for approval. Every tiny thing required ten conversations, all of which led back up to the top. Quickly it became clear that we had a bridge-and-troll situation. The VP was the only one who could approve any website change as low level as correcting broken link repair. There were multiple senior marketing managers who reported to the VP, yet none of them could approve anything with finality, not even a blog post. We made updates that we subsequently saw overwritten by the client on multiple occasions.

We provided endless useful suggestions and tried to mediate what were becoming difficult conversations. When consulting on an SEO project, it’s a 100% great idea for both sides to make sure there’s zero mystery where the time is going. People shouldn’t have to pay for mystery. Sometimes organizations get super busy for a spell. When we eventually got a thumbs-up to proceed on the project’s tasks, we took extra care to make sure it was all documented. There was no way we had their full attention. The VP of marketing was anxious to see results while simultaneously being unwilling to let us do what needed to be done.

We were putting ideas out there to fix problems that we found, like providing conversion-increasing design comps. Our layouts highlighted their product more succinctly. We created stronger calls to action. What we actually managed to get implemented on the homepage showed almost instant organic benefits. We spelled out our strategy, but it wasn’t being heard. The person who brought us in was powerless and eventually every conversation devolved into, “OK, I will talk to _____.” We were dancing around an elaborate system of senseless roadblocks. This was an increasingly untenable situation unlike any I’ve ever encountered in my career.

The content, design updates, analytics, and layout changes we made to the site started to help increase the organic traffic. Then the PR consultant secured a hit in a major news outlet. We did a small social campaign at the same time, which was unrelated. Unfortunately, the PR consultant forgot to ask for links back to the company so none of the press hits contained links to the homepage. We identified that the traffic burst in analytics was from our campaign, not the press.

Because of the press hit, the client refused to believe that the traffic was from us, despite having the analytics in front of them displaying where the traffic came from. We were not hired for PR but seemed to be involved with it, because our influencer contacts ran deeper than the other resources engaged. The constant content and PR discussion led us away from SEO as a primary task. Scope creep is a tangible issue to watch out for in any consulting project and it always takes the search program off track. The general drama running rampant in this organization accounted for much of the business not getting handled, so eventually it became time to move on.

When people coming together on a project are not operating with common goals (in good faith), it’s not worth the risk to continue. It came out a few years later that there was some corruption within the company that led to quite a bit of investigation by press.

Lesson Learned: When there is no remaining move to be made, the search battle is over. The goals have to line up across all relevant parties when you’re building out a search program. It’s normal to have a tiny bit of resistance when you are trying to physically change the innards of a marketing department, but too much resistance is a red flag. The marketing department’s head did not allow so much as a blog post to go live without approval, and we were never going to win.

Social Success!

This was the case of a popular online content site, respected amongst the community of tech and on par with major leading business publications. We helped them to find revenue online in a time when online content publishers were struggling to monetize social. Their industry awards drove a lot of revenue for the company as did their live conferences and local events. This project was ultimately a success because social media drove event registrations and award submissions.

This project started when their existing digital agency took down the Facebook page for violating terms of service by using some automated tool. The tool was supposed to enhance reporting and publishing, but it was also untested. While tools are tremendously important to an SEO’s success, they can also lead to his or her downfall. They were fired for this very public and costly error. A significant amount of the magazine’s traffic came from Facebook. In times of adversity, people often revert back to what they know. They understandably felt like letting anyone else handling social was a bad idea. After this experience, there was very little trust for vendors and that was the first mountain we had to climb. We summited and were the ones who got the page back up.

We didn’t do anything fancy to get the site restored; we simply followed the process for dispute. It took three full days, but we got it up permanently. It’s very tempting when you’re faced with a difficult failure to want to call someone and force a solution. With the larger tools, it’s always more reliable to follow the process they outline, and only escalate if that doesn’t work.

To gain trust, we did the work and earned it. We scoured the web for future reputation issues or failures for performance. We studied their data to identify which marketing activities were working and which were not. While the publication was financially successful, they had extensive brand-leaking all over the web. Nobody was monitoring the name.

A dozen or so groups on LinkedIn had the magazine’s name with different versions of the logo. The landing pages they were using for events did not even have the company site in the logo. All assets built for them by the previous agency were done on the previous agency’s properties, not theirs. They were charging the client an astronomical rate to house this content (bitter divorce).

We stopped the bleeding and migrated all relevant content and registration to the page with a better CMS hitched up. Previously they were using a popular open source CMS system that had few associated fees until you took into account the human cost to maintain it. We switched them to a more robust system that required less management, so staff was freed up to build rather than maintain the site.

Fixing the LinkedIn groups took a little more time. We started by picking the largest group that anyone within the organization still had access to. We reported the groups that were rogue or untended to. They also had a record amount of Google+ pages. It took some time to track down ownership and close them out. The only way to get prominence for the LinkedIn group was to turn on the spigot. We posted every single article that came out on the group and used the extensive email list to invite the selected people to the group. Over the course of a few months, we had over 50,000 followers in the group and it was very clear to anyone searching for them that this was the correct group. They also had a LinkedIn company page, which we updated to have a more official presence.

We waited for their annual announcement of the top rankings for the companies they were researching that year to test out our theory about how to get into the LI headlines. I’d researched and tested with some other posts for another client and happened to figure out the formula. We repeated this for the client. Prior to the creation of LinkedIn Pulse, it was a great feat for publishers to get their links in the LI news headlines. We experimented and figured out the right combination of sharing content from individuals, groups, and private messages. Within the company, we coordinated the sharing of the awards both publicly and privately. The results were that they got featured in the LinkedIn news carousel for an entire day. It’s unfortunate that the news carousel no longer exists; LinkedIn Pulse took over. Big sites always want to host content, not refer it out.

All emails were sent with the social badges, landing pages, and registration pages, too. We tweeted for them at live events and coached staff on how to do this. Eventually social media became a key component of their event strategy and was equal to other forms of traffic on the website. When it was award season again, we drove application signups by utilizing social media ads. With email addresses and retargeting options available for the ads, it was the best value. We also tweeted at individual companies with significant social reach to congratulate them.

Lesson Learned: Never give away your content without benefit. Experimentation leads to new knowledge, which can then be used for traffic gain. Diversified strategies are generally the best; just because something is working now doesn’t mean it will forever. The strongest long-term strategy is to never form a dependency on one lead source. Yes, you can milk it as long as it lasts—but wells tend to run dry. Keep challenging yourself to find new strategies. Always maintain autonomy.