Starting at the end - The campaign's end goal was to drive sufficient site traffic to convert 50+ quality applications to Ordnance Survey's Geovation Programme within three weeks - drawn from an extremely niche audience of technology start-ups.
Ordnance Survey had recently launched its London-based Geospatial Innovation Hub with a view to increasing the adoption and application of its UK geospatial assets among new and existing industry partners. Coupled with this was the launch of their Geovation Programme, an accelerator programme targeted to individuals and start-ups that want to create a product or a business using geographic data representation and analysis. Successful admission to the Programme was dependent, however, on three key requirements - that the candidate individual or start-up:
- Already had a structured idea for a product or business that could make use of OS's geospatial technology;
- Currently had no access to investment or seed capital;
- Were able to spend at least 20 hours per week working on their project and attending core workshops at the Geovation Hub in London.
Naturally this pointed to a very specific audience. When launching the Programme, Ordnance Survey had set a cap of 10 admissions to the Programme. In order to ensure that these admissions were of optimal quality, they set a target of 50 applications from which to cherry-pick the strongest proposals. So defined were the target audience parameters, however, that, by mid-July, four months after launch, the Programme team’s attempts to attract prospects via targeted DM and Social Media, had failed to attract a single application. With only three weeks to go before the application deadline, the Programme team turned to Hunterlodge, with what was undeniably, a time-bound challenge. Budget was extremely limited too - so it was imperative our targeting was ultra-relevant, and that all activity was measurable from day one, to facilitate ongoing optimisation and make those challenging targets attainable.
The magic ingredient
Recognising the need to drive sufficient site traffic that would convert to 50+ quality applications within three weeks, we recommended digital channels and formats that we could get live within 2 working days of receiving the brief. We quickly audited the size of opportunity in each available channel, set up target audiences and ring-fenced the appropriate resource over the first 48 hours to ensure the speediest execution of this critical campaign. Given the tight time-frame and limited budget, we devised a sensibly cautious plan that, for each proposed channel, included 2-3 different targeting strategies. This ensured that we could gather learning quickly, observing which target audiences were engaging most deeply, so that we could nimbly shift budgets to those most responsive.
Highly targeted media selection
For specific channel selection we turned to our digital media insight tool, Global Web Index, to ascertain the profile and media channel preferences of our core target audience â€“ Software Developers. We found that they were mostly Male (88%) with the vast majority (78%) of those harbouring ‘entrepreneurial ambitions’ aged between 25-44. Channel-wise, we found that these developers do access Social Media, but not for professional purposes. We found that they prefer self-help and self-solve, and to discuss technology in open forums. For this purpose LinkedIn was deemed imperative for our successful engagement of them: 75% of developers, we found, access this channel on a weekly basis, because it ‘facilitates learning’ and is ‘easy to search’whether for articles, videos or people.
Heeding this we recommended Sponsored Updates on LinkedIn, editorial links that show up in the targeted users newsfeed. This served both our need to satisfy developers’ hunger for job-related content and the pressure to get activity live as soon as possible. As Ordnance Survey already had content about the Geovation Programme on their LinkedIn page, it was a quick and easy process of simply promoting that content rather than waiting for the creation of premium display assets.
To gather learning quickly we initially targeted these updates to three distinct audiences:
- Greater London-based (so that they could, realistically, attend the London-based hub for 20+ hours a week) Men, aged 25-35, working in software development such as Architect, IT project/programme manager, back-end developer, IT analyst/consultant etc prioritising those conversant in Java, .NET, UNIX of C languages (Targetable volume on LinkedIn: 302,000);
- Greater London-based Men, aged 30d with technology and/or business degrees - again, prioritising those conversant in Java, .NET, UNIX or C languages (Targetable volume on LinkedIn: 72,000);
- Heads of technology-related businesses with 0-5 employees (Targetable volume on LinkedIn: 7,000)
For slightly more broadcast (but still ultra-targeted) reach - essentially, putting the Geovation message in front of secondary audiences (i.e., key influencers) as well as potential applicants themselves, we recommended Programmatically-served Display advertising (testing all formats and optimising by format and message from day one). For optimal performance we recommended running two targeting strategies for this activity and shifting budget to one once we had a clear view of which was driving best engagement and conversion:
- Display targeting strategy 1: serving ad impressions exclusively to Greater London-based professionals aged 25-34 that visit a white-list of cherry-picked technology news/career websites;
- Display targeting strategy 2: serving ad impressions network-wide to Greater London-based Careerists, aged 25-34, working in Technology & Computing (sub-category: Databases, Data Centres, Internet Technology)
We were aware, of course, that our activity on LinkedIn and our targeted Display activity would generate awareness and build consideration that might not be harvested directly by those channels themselves. Many would seek out more information afterwards in a research pattern that would inevitably recommence in Search. We needed, therefore, to underpin our content and display activity with a robust Search strategy. To this end we recommended setting up three separate Google AdWord groups: one relating to Brand search terms (e.g., Ordnance Survey, Geovation Hub), one relating to the specific technology (e.g., geospacial technology) and another relating to micro-SME/start-up support (e.g., seed capital, start-up funding). Naturally each of these ad groups would serve ad text relevant to each of those key areas.
We were confident that our quick wins plan would generate 1.1m targeted impressions over 3 weeks, driving 1,377 visits to the Geovation Programme landing pages, resulting in 54 applications to that programme (a conservative estimate but still 8% over target).
Our speed to market and cautious multi-layered targeting strategy across all channels paid huge dividends and worked to show that our forecasts were very conservative indeed. Day-to-day review of all channel performance revealed that
- On LinkedIn, role-oriented targeting garnered better engagement than degree- or Senior Management-targeting and so our efforts and investments shifted to that group predominantly;
- In Programmatic Display, impressions targeted to specific career categories proved more responsive than those targeted at a whitelist of tech sites, so we shifted out focus accordingly;
- In Search, certain keywords in both technology- and SME-related ad groups were contributing to the lion’s share of conversions so we upped the bidding on those, and dialled up investment in Brand terms, as word of the Programme spread through Social networks.
We also noted after the first week of activity that, by far, the highest engagement levels (i.e., click volumes and number of pages visited) came at weekends. This, we deduced, was a period when our time-poor target audience had more time to digest content and, one would assume, to complete an application. In recognition of this we ring-fenced investment for weekend up-weight to ensure we harvested response from a more-engaged audience.
All of this cautious implementation to gather insight - and our rigorous optimisation based on that learning - worked: before the final weekend the Geovation Programme had received 49 quality applications. After the final weekend, on the final deadline day, that number had reached 77 applications (54% over the initial target) which have since gone on to comfortably deliver the 10 quality admissions to which the Programme aspired. A resounding success.
”This was a complex requirement aimed at a highly niche audience. It required careful insight generation and tailored engagement strategies to drive applications from the right people at the right time. Combine this with the extremely tight time-schedules and the results are even more impressive!”
Gareth Miller, Head of Marketing Communications