Stratasys is an industrial 3D printer manufacturer that also maintains, among other subsidiaries, a parts on demand division called Stratasys Direct.

Customers can choose to buy a 3D printer and manufacture parts with it, or they can upload a computer-aided design (CAD) file to the Stratasys Direct website and have the company 3D print the part for them.

This involved two entirely separate go-to-market strategies: to buy a printer, customers typically utilize Stratasys' global reseller network.

Here's the thing, though: some Stratasys 3D printers can cost as much as $500,000. And then the customer has to buy the materials to feed the machine (like how blades have to be purchased repeatedly for a razor).

At Stratasys Direct, the process is decidedly more straightforward, and much of the business is conducted (or at least started) online through a quoting portal.

For parts with more standardized requirements, eCommerce options are available. For an additional cost, the customer can work with a project engineer to help with selecting the correct technology and material.

For the first 3 years I was at Stratasys, I worked in the Stratasys Direct division. I was hired as Senior Growth Marketing Manager, and was tasked with creating demand during a challenging period for the business.

I had arrived at an interesting time (January 2017). Over the past several years, Stratasys had made a number of acquisitions — so many, in fact, that it was hard to keep track of them all.

These acquisitions were made possible by a massive influx of capital in the form additional shares made available for purchase (an increase of nearly 80% from 2012 to 2013 as a result of a merger) and an inflated stock price.

For those who don't remember, there was a huge amount of hype about the 3D printing industry in the early 2010s. Industry pioneers like Stratasys and 3D Systems that had been around since the 1980s got busy.

In 2011, Stratasys acquired SolidScape. In 2012, Stratasys merged with Objet. In 2013, it acquired MakerBot and its content platform Thingiverse. In 2014, it acquired GrabCAD.

Also in 2014, it acquired Solid Concepts and Harvest Technologies, two parts on demand companies that Stratasys merged with its existing RedEye subsidiary to form Stratasys Direct.

After the latter round of acquisitions were completed in 2014, Stratasys' market capitalization surpassed a then-staggering $6 billion. By early 2017, the stock price had crashed, and the market cap was down to just $1 billion.

3D printing had been around for decades as a solution for B2B applications (especially prototyping), but the media-driven hype that 3D printing would become a B2C solution that would transform everyone's homes fizzled completely.

It can be quite challenging to join a company after so much rapid internal change, to say the least. My job was to come in and create demand for the recently formed Stratasys Direct division.

I spent the first half of 2017 largely rebuilding Stratasys Direct's Google Ads account from the ground up.

When I took a look at which keywords were converting to pipeline and revenue in our Salesforce CRM, I quickly saw a problem: almost all of the revenue was attributed to keywords that contained the company brand.

Branded keywords are cheap; Google charges almost nothing for users to click on those ads if it's your brand since you're providing the user with a good experience. (As general rule, if the user is happy, Google is happy.)

This meant, however, that almost all Google Ads spend was being wasted; very little of the revenue was coming from the keywords that were costing us a lot of money.

After almost fainting when I realized the full scope of the problem, I took some deep breaths and went about entirely rebuilding the account, from keyword selection to campaign setup to ad copy to landing pages.

Once this was accomplished and campaigns were pushed live, the real work began. For instance, negative keywords needed to be added to prevent ads from showing up on searches we didn't want.

Having to immediately go to work fixing our Google Ads account turned out to be invaluable, since this directly informed the SEO strategy I had to put together for the website migration project in the second half of 2017.

For instance, our website at the time was very focused on the various 3D printing technologies we offered, and only mentioned materials in the context of the technologies (e.g., materials for FDM).

Since 2013, Google's policy of encrypting search data has resulted in organic keywords being "(not provided)" in their reporting. However, the Search Terms report in Google Ads shows precisely what the user typed in.

Why did Google remove the ability to view actual organic search keyword data? Because they want you to have to spend money on their search advertising of course!

The conversion data in the Search Terms report turned out to be really interesting in two key areas, in the sense that they revealed the thought process of potential customers in ways I hadn't considered.

The first was that people often search geographically when they don't know where to start. Did we have an office in New York? No, but plenty of people typed in "new york 3d printing services" and converted on our website anyway.

To be clear, we were not bidding on any geographic keywords at the time. But when we bid on, say, the phrase match of "3d printing services", our ad would show when someone typed "new york 3d printing services" into Google.

And since much of our business was conducted online or over the phone, it really didn't matter where people were coming from. The logical next step was to target US states and cities with our keyword research and move in for the kill, which is what we did.

Another surprise that became critical in informing our eventual SEO strategy later on was that many conversions on our website were for search terms like "abs 3d printing" and "nylon 3d printing".

In other words, these users had the material in mind first before any particular 3D printing technology (nylon 12, for instance, can be 3D printed using one of four technologies: FDM, SLS, MJF or SAF).

However, the website (and our paid search campaigns) at the time only presented it in reverse: pick the technology, then see which materials were available for that technology.

Some users, of course, approached it from that angle, but there was an entire group of users who approached it from the opposite angle — and nothing on the website offered them a customer journey that aligned with this.

What's more, there was a lot of search traffic for keywords like "abs 3d printing" and "nylon 3d printing", not to mention there was an absolute ton of traffic for materials in general.

However, there were no pages on the website that could properly capture this traffic — for example, a page dedicated to nylon did not exist. An entire materials section needed to be created.

As it turned out, a proper materials section was not built out for the website for another year. Before any of these major SEO gaps could be addressed, the website needed to be migrated from WordPress to Sitecore.

When Solid Concepts and Harvest Technologies were acquired by Stratasys in 2014 and merged with RedEye On Demand to form Stratasys Direct, the content on the three previous websites was merged together as well.

A new website was created using WordPress. WordPress is frequently the content management system (CMS) of choice for businesses in their earliest stages due to its flexibility, vast plugin library, and relative ease-of-use.

It's easy to get a WordPress website up and running quickly. But when I arrived in 2017 the website had already existed for nearly three years and my colleagues across the company were itching for a better site (they weren't subtle about it).

And for good reason: a business' website is its number one sales representative. Obviously, a website has to convey the company's branding and messaging, but above all it needs to create demand.

We essentially had two choices: work with an agency to design us a new WordPress theme that was completely customized to our specifications, or be brought onto the Sitecore environment that housed Stratasys.com.

Sticking with WordPress was by far the easier option. I was intimately familiar with WordPress and also had just spent several years at agencies. I knew exactly what to ask for and how to avoid being overcharged.

But as is so often the case in Corporate America, decisions like this are made at by people many levels above you in the organizational chart. The choice was made to bring StratasysDirect.com onto Sitecore.

From an organizational perspective, this decision made sense: the websites would now look alike, and Stratasys was keen on more closely integrating Stratasys Direct into its go-to-market by having it be the "tip of the spear" for customers who might invest in a printer later.

The enormous challenge, of course, was that I had never used Sitecore before, and neither had anyone else on the Stratasys Direct team. And unlike WordPress, Sitecore is not easy to learn.

Thankfully, our pals at Stratasys marketing in the American corporate headquarters in Eden Prairie, Minnesota could not have been more gracious and willing to bring us up to speed.

In fact, we coordinated across not just American time zones (I was based in the former Solid Concepts headquarters in Valencia, California), but internationally, as well. Significant development resources were provided by the former Objet headquarters in Rehovot, Israel.

Familiarizing ourselves with Sitecore turned out to be an ongoing challenge — even after the migration was done, we still were figuring out new things we could do with it.

Essentially, we were provided with custom-developed components in Sitecore that we could configure to build pages with. Components are like building blocks — you can move them around to shift sections higher or lower.

We had to take these components that had been specifically developed for the Stratasys 3D printer business and figure out how to create pages that would fit our specific needs for the parts on demand side of the business.

It was hard — by no means was this a simple lift and shift. This was a ground-up rebuild using a new CMS, with custom-developed components that were originally designed with a different website in mind.

This meant taking the existing content that was on WordPress and figuring out which Sitecore components would be best at presenting it. This required a lot of vision, and we had to do it for hundreds of pages.

Then the content had to be moved over. First the components were built out, then the components were added to the page, and then the page was published. Over and over (and over) again — it was mind-numbing.

Thankfully, all the work that had gone into the SEO strategy months earlier paid dividends. The URL structure had been cleaned up, the pages re-titled and positioned to go after the most possible search traffic, etc.

I will say, there are times when I am in the middle of these types of gargantuan projects and I wonder whether it will all be worth it after spending six months in the dark, not being able to see any results yet.

This is just the nature of the beast, and why it's best to avoid these kinds of prolonged development cycles in the first place if you can help it.

Sadly, the only way out was through in this particular case — with two different CMSs in play, we couldn't selectively migrate over and push live a few pages at a time in a structured, iterative process.

Although the new website was live, we still had to wait a little while for results on the organic search front. In the meantime, there were small things to clean up (as there always are immediately after going live).

In particular, we needed to make sure there weren't any leaks in the conversion funnels and that our analytics data was clean. Google rankings were important, but user experience was paramount after the go-live.

Sure enough, as days turned into weeks and weeks turned into months, we began to see rankings shift, traffic increase, and conversions elevate.

There are three ways to increase conversions on a website. The first is to increase the amount of traffic. The second is to increase the rate at which the traffic converts. The third is to do both.

Of course, those terms are used quite broadly above and don't convey the entire truth (though the point is still valid). Certain conversions are better than other conversions, and some traffic is better than other traffic.

This is illustrated by the buyer's journey, which consists of three stages: Awareness, Consideration, and Decision. Awareness = top of funnel, Consideration = middle of funnel, and Decision = bottom of funnel.

Paid search allowed us to maximize our presence in front of Decision-stage traffic, thanks to the ability to bid on transactional and commercial keywords.

We of course wanted to get Decision-stage traffic organically as well, but generally speaking the kinds of landing pages built for paid search advertising do not rank organically. (The best practice is to noindex them.)

Paid search landing pages convert well by not overwhelming the user with information. The mission is singular: convert the user to a single goal before they click on the back button.

By contrast, pages on a website that rank well organically tend to be highly informational and detailed. Typically, this type of content is good at helping users who are becoming aware of the problem they have.

For Awareness-stage traffic, the goal is usually to get them to progress through the website to Consideration-stage pages, where they can convert for a middle-of-funnel conversion (like downloading a design guide).

Of course, the ultimate goal is for users to sign up for the quoting portal and create pipeline and revenue, but the simple fact is that most website traffic just isn't at the Decision stage — you have to provide other user journeys.

A year is a long time to be able to gather data and continuously iterate. By looking at the Google Ads Search Terms Report, we were able to keep seeing exactly what users were typing into Google prior to converting.

We also could see which pages on the website were not only generating the most traffic in Google Analytics, but which were generating the most conversions. All you have to do is follow the data, and implement changes quickly.

We spent an entire year sculpting a true demand generation system: over the course of 2018 we 3x'd our Google Ads spend from $30k to $90k per month while increasing pipeline from Google Ads by 4.5x compared to 2017.

And as for organic search? We doubled the number of orders from SEO year over year. 2018 was a good year for Stratasys Direct — after several years of industry turmoil and hype unfulfilled, we had laid a critical foundation for long-term growth.