In 1884 Sarah Winchester began building the Winchester Mystery House, a mansion with doors opening nowhere and stairs ascending into the ceiling across an estate spanning 160 rooms, 6 kitchens and more oddities than any other property of its era. Eschewing the services of an architect, she changed plans daily and had to pay her carpenters a 50% premium to tolerate the frustration. Measuring today’s multichannel marketing efforts can feel like building a Winchester Mystery House for marketing – data sources over here, campaign reporting over there with no unifying plan.
Addressing the problem starts with an architecture that accommodates both automated and non-automated systems, notes Martin Kihn, research director for Gartner for Marketing Leaders. The architecture has three parts, starting with great key performance indicators (KPIs).
1. Define metrics and identify data sources
Begin by setting customer engagement KPIs, such as campaign channel response rates or revenue from lead gen activities. It helps to start with embedded off-the-shelf metrics that take advantage of automated marketing platforms and move into more advanced, strategic metrics that might require manual analysis. Once the KPIs are set, the next step is to integrate data sources from creative agencies, marketing systems, campaign reports and other sources which may require database normalization and management capabilities.
2. Build revenue and cost models to calculate ROI
The next area to tackle is attributing revenue to various channels. Channel attribution analysis aims at the revenue or business benefit contributed by particular channels or media within a campaign and can be gathered from within various marketing or campaign tools. Costs are then calculated for the various channels, campaigns or programs, including media and related overhead. Analyzing cost vs. revenue for a particular campaign or program will deliver the cross-channel ROI figure.
3. Design reports and dashboards
The key to great dashboards is designing for specific roles and including only the “critical few” measures that matter to that role. For example, the CMO dashboard should contain simple KPIs that describe the past (what happened), diagnose the present (big changes) and look forward (for example, pipeline). They should cover key areas of the business like financials, voice-of-customer, operations and campaigns. Alerts can then be set to indicate progress against goal or previous period and variance from thresholds. Finding the right vendor for reporting means considering a mix of service and software providers and in-house solutions; many marketing teams still have to build parts of their solution in house to meet unique needs.
In summary, leverage an architecture that includes data integration, business intelligence, advanced analytical tools for attribution and optimization, ROI analysis, role-based dashboards, and alerts/ notifications to avoid building a cross-channel marketing mystery house.
