Published by World Advertising Research Center Farm Road, Henley-on-Thames, Oxon RG9 1EJ, UK Tel: +44 (0)1491 411000  Web: www.warc.com

November 2006, Issue 477

The discipline of ROMI (return on marketing investment) has enjoyed tremendous success over the last five years. It is now in widespread use across companies not just in consumer products but also in industries such as financial services, automotive, communications and technology, and even government agencies.

The appeal of ROMI models is easy to grasp given today's marketing environment. Executives face increasing pressure to drive profitable growth while having to justify investments in driving this growth using the languages of capital budgeting and strategic planning spoken at the chief financial officer's office.

This is both good news and bad news from the perspective of migrating towards a more comprehensive understanding of ROMI. The good news is that the idea and mindset of ROMI are well engrained and lots of historical experience and benchmark data are available. The bad news is that current approaches to ROMI typically offer an incomplete view of how marketing drives value, and do not provide an ideal platform for integration into the realities of marketing planning.

In recent work for a client in the relatively stable CPG sector, we compared using simple marketing-mix model (MMM) results to predict in-market results, and compared these with a developed forward-looking methodology for investment allocation. In this live experiment, the mix model had a forecast error of 13% compared with the more detailed, integrated approach which had only a 3% error range. Why?

ROMI analysis in general, and MMMs specifically, use historical data to measure how advertising and marketing worked in the past. Perhaps it should not surprise us that these reports tend to be used mainly for historical validation and not, as they promise, for forward planning. There are three main reasons why market results differ from these historical 'forecasts'.

1. Predictability. they have a low level of predictability in rapidly changing environments – for example, re-branding and innovations (both product and media) have no historical data.

2. Granularity. the models are not specified with enough detail to capture media planning realities – for example, reach, frequency, recency, relevance and synergy.

3. Relevance. the models do not capture the full impact of marketing programmes, for example, long-term effects, impact on pricing power, brand-extension opportunities.

Essentially there is a disconnect between ROMI analysis and planning because it is too high-level and historically focused compared to the more detailed, innovative and forward-looking nature of planning. Further, historical applications of ROMI models have been focused on understanding the short-term response to marketing investments, the primary use being to identify least-performing programmes that can be cut under budget rationalisation. When this is done without full measurement of a programme's impact, it may lead to worse long-term performance as demand- and profitability-generating investments are reduced (see discussion Box).

So, both for planning and for growth, a broader understanding and model for managing marketing effectiveness is required.

CHANGING EFFECTIVENESS AND INNOVATION

Marketing effectiveness is affected by many more factors than a ROMI model typically measures (see box, right). As these factors change through time they surely impact expected effectiveness. So, for example, just because a radio campaign had a low ROI last year doesn't mean that, with the new product and channel partner in place next year, it will not then have a high ROI. To be fair to MMM, certain rules of thumb have emerged from the industry's experience: it is accepted that media produce higher returns (sales impact) when associated with a new product introduction or brand innovation. Again, sales-response curves typically have a similar shape (ramp-up and saturation) as communication vehicles reach curve. TV usually has the largest potential sales impact, though other channels may prove more efficient at lower spend levels.

Such general observations are reassuring, but what is needed is a much more formal, complete system for integrating all the factors that can influence expected effectiveness, to provide for a more predictive investment – optimisation model.

Now that many ROMI models are in place within companies, and have been for a while, secondary analysis (bench-marking) is possible to measure the impact of these factors on changing effectiveness. Coupled with the formal integration of insights and research from client and agency teams we can build a process and toolset for managing and predicting these changes and their impact on future programme ROI.

A recent project required us to build the case for significant budget increases – more than double – given aggressive sales targets for a new model line. Historical analysis could not support the case, so a broader methodology was used to capture expected sales under the innovation and new communication opportunities that the higher budget would enable. The budget was awarded and the sales goals are being met.

BETTER THE DEVIL YOU KNOW?

Aligning ROMI with the planning process is also an issue of cultural adoption. The emphasis of MMM has previously been on short-term sales impact, while many communication innovations are focused on more subtle consumer and brand objectives. This has tended to make MMM the pariah of innovators. To achieve harmony and agreement here, we need to integrate several complementary and corresponding research sources to enable the valuation of new media opportunities alongside traditional ones. We also need to place historical results in the context of today's communication environment, and the expected environment tomorrow.

That said, we still need a fairly dispassionate method for distributing funds, based on maximising value for the business.

Some innovations are essentially risky investments, but with potentially higher returns. MMM represents a known, though potentially lower, return, and in an environment of short-term shareholder pressure, risks are made to look very unappealing. What is needed is a forward-looking and financially robust method to compare investment options, to maximise long-term business performance and shareholder value.

An obvious structure for integrating communications, innovation and sales impact is the customer funnel. This enables us to marry communication and marketing metrics with business financials through a set of properly aligned KPIs: not just something for everyone, but a systematic and complete approach to mapping marketing value creation at the customer (top-line) level.

TOTAL VALUE CREATION AND OPTIMISATION

With some modification (see Figure 1), looking through the lens of the customer funnel provides an approach for measuring more than just short-term sales response. There are five other areas that we typically look to for additional value creation:

1. building pricing power by reducing price elasticity – which increases promotional efficiency

2. building brand equity to enable successful line extensions or licensing opportunities

3. building greater programme and organisational alignment between demand-generating groups and sales management teams – creating synergies that drive greater acquisition/trial conversion

4. building long-term or latent demand that may materialise as consumers come into the purchase cycle

5. creating goodwill and traffic with distribution and channel partners.

In terms of the funnel, increased pricing power, channel strength, and long-term repeat purchase all result in a widening of the funnel and increased velocity in customer flow from acquisition onwards. In terms of long-term latent demand under longer purchase cycles we see widening of the funnel at pre-acquisition stages.

The two other value-creation areas require additional structure to the funnel. Building synergy between marketing and sales focuses attention on customer activation. Sitting between acquisition and retention, recognising this stage of activation enables us to incorporate shopper marketing, ERM solutions and other 'early customer' marketing programmes that are vital at the initial stages of customer-relationship creation. Brand equity that supports innovation and licensing leads us to identify an ultimate stage of optimisation. Here we have probabilised returns from traditional cross-sell and up-sell opportunities as well as future innovations and extensions.

Mike Campbell provides an excellent discussion of the current state of play in measuring some of these additional value-creation sources (1). As new statistical techniques become commercially available from the academic community we can produce more reliable models to measure some of these other effects.

From an optimisation perspective, excluding these additional returns, which conceivably vary across different marketing programme channels, will lead to misallocations in mix, and under-allocations in total budgets. Compounding this issue is the general use of MMM to weed out under-performing programmes under budget restrictions, not for total-profit maximisation.

An example of the benefits of including these additional returns was seen in a recent financial services case. The process identified changes to budget-setting and allocation that increased returns on the marketing budget from 20% to 45% through a combination of scale deployment of new channels, budget increases and a rebalance of focus between acquisition and retention.

CONCLUSION

In short, an approach to investment support for profitable, value-creating growth is required. Such an approach should build on the insights provided by historical analysis, but clearly needs to go beyond it to be of real value to both business and communication planners. Integrating MMM and other insights such as panel data-switching models, direct-marketing post-campaign analysis, segmentation and brand audits, into a forward-looking, predictive optimisation model requires some orchestration.

Methodologies need to be built that address each of the areas that affect marketing effectiveness (see Figure 1) and relate them to the expected sales response of future channel use. Then forecasting these change factors themselves requires deep insight that often only resides in companies dedicated in those areas, and that often means the marketing channel agencies. This is clearly most acute when we consider integrating new and previously unused media channels, which continue to proliferate.

That said, it is not just a case of clever analytics and models that will address the need here. It also requires process and workflow changes to institutionalise both the translation of various historical analyses to future predictions and the use of these in channel planning and budget allocation. Can such an investment in methodology and process change pay off?

The benefits are twofold, with improvements in both business financial performance and in process efficiency (see Figure 2).

Taking a more financial approach to supporting innovative marketing opportunities, as well as better identification of the total returns to all marketing programmes, can yield substantially greater returns. In our experience working with clients to support growth programmes and more holistic investment analysis, additional returns well in excess of 25% are achievable.

Marketing-mix models still have their (valuable) place, but can be seen only as part of the process and an input to a richer methodology for investment optimisation under a broader model of marketing effectiveness.

1. M Campbell. Is ROI Dead? Admap, March 2005, Issue 459.

APPENDIX

ROI – Finding the Right Metric(s)

As a result of ROMI-based budget rationalization and the drive for short-term shareholder values consumer products companies in the US have over the last five years experienced unprecedented reductions in marketing budgets expressed as % of sales (see Figure A).

Senior executives striving to integrate marketing investments into capital budgeting and strategic planning frameworks could argue that this is a result of understanding the actual contribution and adjusting spends to maximise returns on invested marketing capital. Looking at returns for US consumer products companies over a similar periods one would be tempted to conclude that the strategy of maximising short-term returns has worked (see Figure B).

Returns on invested capital (ROIC) have generally increased in the industry over the five-year period.

Meanwhile, there are two major issues with short-term ROI as a performance metric in general and for marketing specifically:

1. Maximizing ROI versus maximizing value

The notion of maximizing ROI would dictate that resources be allocated so as to generate the highest possible return on that investment. The problem with this is that it would throw out programmes that are profitable if they happen to have the lowest ROI estimate. While this is the most efficient use of a restricted amount of funds, it does not maximise overall value creation by the firm. This is a key point – the objective should not be to maximise short-term ROI, but long-term absolute value creation in excess of the cost of capital (see Figure C).

2. Capturing the entire picture

The second major issue relates specifically to marketing. First, when constructing a ROMI framework as input to the planning process, this must capture the true universe of investment options available, and capture them in a way that reflects their expected future performance. Typically, traditional ROMI models focus only on short-term volume response and ignore some important ways in which marketing drives value. Failure to quantify these other effects, such as increasing pricing power and brand-extension opportunities, leads to over-reliance on short-term measures and an inability to demonstrate acceptable returns from activities – and thus to reductions in budgets.

Do short-term returns maximise long-term value?

This brings us back to the reduced budgets and increasing returns in the US consumer products industry. As shown, evidence would suggest that the strategy has paid off. But as we decompose the metric, we see that absolute sales and profit growth has declined over the period, while short-term relative returns have increased (see Figure D).

Does the Street reward short-term behaviour?

Short-term ROI maximisation has frequently been justified to meet the demands of Wall Street. Meanwhile, multiples across US consumer products companies have been declining even as relative ROIC has increased. This means that long-term growth and profit remain true value drivers, so marketing must find comprehensive ways to influence, measure and manage these (see Figure E).

http://www.warc.com

© World Advertising Research Center