data-driven lead generation & product launch for control bionics
🏆 Winner of the 2019 AMA Cincinnati Research Insight Catalyst Award.
All data and visuals shown are publicly available. Audiences and some data have been obscured to avoid revealing information proprietary to the client.
Opportunity / Problem
Control Bionics developed and sells NeuroNode, an advanced communication device for people with ALS (Lou Gehrig’s disease), spinal muscular atrophy, and other conditions with paralysis and loss of speech. The company was founded by Australian CNN news anchor Peter Ford, and in 2014 partnered with Cincinnati-based TAI to form Control Bionics.
Devices from other companies use eye tracking to control a tablet. Eye tracking is relatively quick but requires a lot of concentration and can be tiring. NeuroNode instead is a device that sits on top of the skin and detects faint electrical signals from the user’s nervous system, present even with paralysis, allowing them to control a tablet. Technically the system is referred to as EMG – electromyography. In spite of the fancy name, it’s quicker to learn and less tiring than eye tracking.
Control Bionics had perfected the device and manufacturing process a year ago and had the capacity for rapid growth. They were running campaigns with another agency using Google search, display, and video, and Facebook posts and ads when they came to us in August 2018, unable to break out of their previous lead numbers.
Primary objectives were to…
Increase qualified leads, not just incrementally but by multi-100%s
Accomplish this in four months, including research and testing (Sept-Dec 2018)
Additional objectives that developed…
Test language and visuals for the benefit of all their marketing
Generate word-of-mouth buzz among clinicians and users in the augmentative and alternative
communication (AAC) community
Promote the product launch for NeuroNode 3 at the ATIA convention in Jan 2019. The new device
combined EMG and eye-tracking into one system, and they wanted a big splash.
PLANNING & EXECUTION
We use Facebook + Instagram as the world’s largest focus group. It’s an inexpensive big data source, and you can segue directly from testing and validation into productive advertising.
By way of comparison, search and display advertising also allow testing, but they only give you binary click/no-click results. With social you get likes or reactions, comments, shares, direct messages, read-more clicks, view-image or -video clicks, and outbound clicks – information about how the information is being received, like a light-weight focus group. Social media can also display your content in a rich set of formats – static visuals, videos, headlines, supporting sentences, as well as complete landing pages, and these can be presented in untidy news feeds, immersive stories, inline article ads, annoying interstitials, text-only teasers, and messenger ads.
The first month of every campaign is devoted to research and setting up the ads. The second month is the first month of actually running ads – this is devoted to testing and we don’t expect results to move the needle. Significant results start coming in the third and fourth months.
google search advertising
When Control Bionics first approached us, their goal was to ramp up leads on Google Search. They figured that their nearly non-existent Facebook results meant that channel couldn’t be productive. We showed them that they were already saturating Google, just spending too much for it, and assured them that Facebook + Instagram had enormous potential for them. The plan that resulted was to make Google simply more efficient, with incremental increases if possible, and divert those savings into Facebook + Instagram, where we expected major growth.
Google Search performed worse than expected out of the gate due to unproductive keywords that we’d added – lots of impressions, but poor click-through rates and even poorer conversions. Yet immediately we also started getting valuable information. We experimented heavily with keywords, including broad-, phrase- and exact-matched and negative keywords. The wording situation for their products is described in
the Phraseology section below. We also experimented with how the keywords were grouped, and spending allocations for ad groups and countries (US and Australia). These experiments continued for several months.
There are many different conditions with paralysis and/or loss of speech; NeuroNode and its eye-tracking competitors are beneficial for certain combinations. Nearly everyone with ALS (Lou Gehrig’s Disease) or spinal muscular atrophy will benefit from one of these devices. Some people with cerebral palsy, and people with certain types of spinal injuries and a variety of other conditions can also benefit.
Stroke is a common condition with paralysis and loss of speech that isn’t a good candidate, and people use and respond to many different words for stroke-related needs. On Google Search these can’t practically be filtered using negative keywords without blocking many appropriate searches, and in advertising we had to develop the right wording. When our language was too broad – in search keywords and ads, social posts and ads, and landing pages – Control Bionics received leads from people who wouldn’t benefit; when we tightened and only mentioned ALS and SMA, they lost out on others who would benefit. This testing went for several months until final language was settled.
In this report we’ve refrained from showing actual data, keywords, and audiences, and instead have relied on verbal descriptions, to avoid disclosing information confidential to Control Bionics. These live sources and the supporting material below show publicly-available information in actual use.
Facebook + Instagram advertising
We created a variety of on-page posts and off-page ads aimed at likers, website visitors, and numerous particular audiences. On-page posts were boosted to likers and other selected audiences for word-of-mouth buzz. For leads, we used off-page ads.
We started with single-interest audiences: ALS, spinal muscular atrophy, assistive technology, speech language pathology, etc. After testing these and refining images, we started combining interests. One audience was people interested in both ALS and assistive technology, another was both ALS and speech language pathology – the combinations further qualified people who saw the ads. We kept things fresh by adjusting spending from time to time to keep repetitions reasonable and using a wide variety of images to be shown.
Like Google Search, FB + IG performed worse than expected out of the gate. Which is to be expected. On social it was due to a combination of text-heavy initial landing page, visuals, text and audiences. Everything basically, since we were just starting testing.
Altogether we tested 44 different audiences and, combined with about a dozen visuals, ran nearly 300 individual ads before settling on 15 audiences and over 100 ads in the US and Australia.
remarketing to website visitors
We remarketed to their website visitors from Google Search on Facebook + Instagram. This marketing wasn’t aimed at lead generation per se, but at giving people a more complete and personal picture of the company’s products and happy customers – both users and professionals – on a light drip basis.
Videos got high engagement and clickthroughs, but middling form completion. We suspect people who love the human interest angle were drawn to the videos, but many didn’t have the same level of immediate need for the device. So we kept videos in the word-of-mouth audiences but avoided them in the lead-generating audiences.
Control Bionics’ existing landing page performed poorly for visitors from Facebook ads. We believe it’s because they’re expecting more visuals and less detail than Google searchers, who are devoting time to sit down and search by words to get information they crave. Social users aren’t looking for this, we’re interrupting them. So we created a new, more visual landing page which performed better for social. See the Live Sources box.
After successful lead growth, the biggest payoff was yet to come. NeuroNode 3’s product launch was scheduled for the evening of Jan 30th at the ATIA Orlando trade show, and by this time we had excellent information about audiences. One week prior to launch, we targeted professionals in the US, using audiences previously developed, on social media with teaser content.
Two days before the trade show, we set up audiences to target ATIA Orlando attendees at the show. We chose interest combinations that performed well in the ongoing campaigns and who, according to Facebook’s tracking, were visiting Orlando on Jan 30th-Feb 1st.
The previous months’ experience was crucial because there’s not enough to time test and adjust in three days, and an audience of several thousand people is too small to get meaningful statistics anyway. The one parameter we adjusted during the short campaign was spend on each visual in order to keep the repetition rate even, since we wanted attendees to see all the ads. We checked and adjusted spend every 4-6 waking hours during that time.
The trade show targeting was done without geofencing, which would only get people once they’re near the booth, whereas we reached them in their hotel and any time they were on FB + IG during the trade show. The ease of remote targeting makes this inexpensive to execute – we were able to provide this service within the monthly advertising retainer the client was paying us already.
GROWTH IN LEADS BY MULTI-100%S WITHIN FOUR MONTHS. This was the primary objective and was accomplished. The first month of testing / advertising – Oct 2018 – actually saw a decrease in leads, likely due to our inefficient allocation during early testing. Following repeated reworking in Oct and Nov, the second month brought the first measurable gains, and by the third month Dec the campaign was firing on all cylinders. This continued into Jan 2019 and beyond.
BETTER THAN 50% LOWER COST PER LEAD. The original goal was to make search more cost-efficient in order to fund social without increasing total spend. This was accomplished. In addition, the cost per lead on social was dramatically lower for them. Initial costs during testing the first two months were higher, but by Jan their monthly ad spend was back to its previous level while sustaining multi-100%s higher leads.
CREATE A BUZZ. Word-of-mouth buzz from users and clinicians in the augmentative and alternative communication (AAC) community. They began to notice this around the fifth month (Jan 2019). There’s no tracking for this like there is for formal leads.
INSIGHTS ON COLORS, IMAGES, PHOTO STYLES, AND LANGUAGE FOR PARALYSIS AND LOSS OF SPEECH SITUATIONS. Nearly everyone with ALS or spinal muscular atrophy can benefit from this device, but there’s also a long tail of other conditions, and yet even more conditions for which it’s not useful. After some testing we worked out the right language to target the long tail without getting the whole dragon.
Control Bionics also obtained insights on what various colors – both in-brand and supplemental – signify to their various target audiences. And which image and photo styles get the best responses. All of these are useful for all their marketing efforts and have resulted in new website banners, new wording, and confirmation of color choices.
LOTS OF ATTENTION AT THE NEURONODE 3 PRODUCT LAUNCH. Control Bionics had non-stop foot traffic from people already interested in the new product to keep three staff people busy all three days of the ATIA trade show.
Even without access to expensive big data, it’s possible to test and validate marketing strategies and advertising decisions. We used this to Control Bionics advantage in 2018, and the campaign continues to run as of this writing. For the period in this report, Sept 2018 – Jan 2019, Control Bionics benefited from:
• Growth in leads by multi-100%s.
• Accomplished in four months, including research and testing.
• Cost per lead reduced to well below 50% of previous cost.
• Word-of-mouth buzz from users and clinicians in the augmentative and alternative communication (AAC) community.
• Insights on colors and image styles for different audiences.
• Insights on language to make the paralysis and loss of speech situation clearer.
• A new product launch with uninterrupted foot traffic from interested parties, occupying three people full time at their booth during the ATIA Orlando trade show.
Social advertising platforms are basically big data vendors in disguise, and we basically use Facebook + Instagram as the world’s largest focus group. The rich set of available audiences, display modes, and feedback types make the platform unparalleled for testing and advertising. The quantitative support and uninterrupted transition from testing to advertising makes a difference for organizations conscious of their budgets and wishing to move quickly.