How an American QSR improved its Marketing ROI by 3x on Food Apps

21st Sep, 2021

 

By Abhinav

Traditionally, Quick Service Restaurants (QSRs) such as McDonald’s or Burger King, have been strategically operating on a brick and mortar model. However, according to some studies, an average QSR generates as much as 75% of its sales from online orders.

With the advent of delivery apps such as Uber Eats and Doordash, a significant portion of QSRs’ business has moved to these platforms. The war to top rank on one of these platforms is an even greater feat. With each brand competing for the top listing, it’s much less about the dollars you pay and much more about optimizing your investments.

The relationship between QSR chains and food delivery apps has its advantages and disadvantages. One of the critical grouses QSRs have against food apps is the incremental marketing spend required to participate on the platform and the inability to measure the impact of their investment. What makes matters worse is the limitation in metrics even available to measure the impact – neither the food apps provide them, nor does anyone else.

At DataWeave, we have made it our mission to enable QSRs to not only define measurable metrics to achieve a positive ROI for food app marketing investments, but we also equip QSRs with the tools to track their competitive performance at granular, zip code-based level so that localized strategies can be modified as needed. Below is an example of a 1000+ store chain QSR we partnered with to optimize a pre-existing investment made with a large food aggregator app. Within months of engagement with us, they were able to achieve a 3X increase in sales without adding any additional marketing dollars.

Below are the pain points we identified and solved together:

1. No Defined Metric

Problem – No leading metric to track marketing performance

One of the first issues we realized was that sales was not a good metric for tracking marketing performance as it’s a lagging metric and doesn’t capture the issues that help grow or suppress sales.

Most of the sales are driven by rank in the cuisine category and searches for branded keywords. But, the QSR chain had no way to track these ranks.

In fact, 70%+ sales go to the first five restaurants for the category and keyword

Comparing ranking on food delivery platforms
Comparing ranking on food delivery platforms across different categories and times

Solution – Establish ranking as a clear marketing metric

By aggregating data across different food app platforms comprehensively, i.e. across locations, at different times of the day, we established the ranking of the QSR chain in critical categories and for priority keywords, identifying where they under or over-performed relative to the competition. As we did this daily- this became a straightforward metric that helped establish the performance of their marketing campaign.

2. Geographical & Categorical Challenges

Problem: Identifying poor-performing stores and zip codes

We realized  it was not a simple exercise to identify well performing stores on food apps since sales depend on many factors such as competition, population of the area, local cuisine preference, etc.

Solution: Zip Code Ranking and Attributes

We tracked the ranking of each store within each Zip Code for keywords and created a list of poor-performing stores. We also extracted attributes such as estimated time of arrival (ETAs), Delivery Fee, Ratings, Reviews, etc., for each of these poor performing stores, to identify the reasons for the poor ranking. 

Analysing key metrics at a store level
Analysing key metrics at a store level – identifying worst & best performing stores

E.g., We realized 356 of the stores were not populating on first page results, primarily because of poor ratings and High ETAs. After the focused initiative, 278 of these stores started showing on the first page and increased sales by 23%. 

3. Sensitivity Analysis Deficiency

Problem: Not clear about the contribution of Rating, ETAs, Fees, etc. on the Ranking

The exact ranking algorithms of these food apps are not publicly shared – so the QSR chain wasn’t clear which variable of rating, ETAs, fees, ad spend, or availability contributed more or less to the overall ranking. 

Solution: Sensitivity analysis for measuring contribution 

Comprehensive data for multiple zip codes in various timestamps was analyzed to determine which variable contributes most significantly to the rankings and when. We also conducted A/B testing – simultaneously testing two different variables, such as reducing ETAs at one store and improving ad spend at another, calculating which led to greater rank and sales impact.

For example, we realized reducing publicized ETA’s (even by decreasing the delivery radius) contributed much more to improve the rankings than changes to ratings.

4. An Unknown Competitive Landscape

Problem: Tracking competitor performance

For example, we found the QSR chain performed well in key urban centers, but the competition was doing even better, but there wasn’t a good way to track and compare the performance of the competitors.

Solution:

We started tracking the QSR chain and the competition for each of the metrics and started comparing performance.

Analysing competitive performance
Analysing competitive performance on key metrics such as ETA, Availability etc

We quickly realized ranking started quickly improving as we gained a slight edge in each metric against the competitors. For example, 5 minutes less ETA adds to higher ranking.

In six months of this exercise with the QSR chain, we improved the average ranking from 24 to 11 for the QSR chain, getting them featured on the first page.

5. Blind Advertising Investment Opportunities

Problem: 

The QSR chain was not clear on which banners (Popular near you, National Favorites, etc.)  to choose to invest in, and had to depend on the recommendations of the food platforms entirely. 

They weren’t even provided a clear view of which position made the banner visible and at what rank among those banners was their promo visible. They were at times the 7th promo in the 6th banner, which has almost zero probability of being discovered by the user – this happened despite paying heavily for the banners.

Solution: 

We aggregated data for all banners populated within each zip code and found out the ranking and in which position the QSR chain was visible.

Analysing right banners
Identifying and analysing right banners for advertising spends

The QSR chain invested in 630 zip code-based banners with guaranteed visibility, but our assessment indicated the banners were only visible in 301 zip codes. After selecting suitable banners for promotions, we improved visibility to 533 zip codes within enhancing the budget.  

We are now using the same strategy for refining discounts, offers, promotions, and coupons. 

6. Lack of Campaign Performance Monitoring

Problem: Unsure of the long-term impact of marketing spend

In general, increasing marketing spend does give a temporary boost to sales, but the QSR chain’s question was, how can we measure the long-term impact i.e., ranking keywords and the targeted zip codes.

Solution: 

We created a simple widget for every marketing campaign which showed the rank for the keywords for selected zip codes before the campaign, during the campaign, and post the campaign, clearly establishing the midterm impact of the campaign. This constant monitoring allowed the QSR to also quickly pivot on their strategy on account of national holidays etc, and act accordingly.

7. Non-Existent ROI Measurement

Problem: Establishing the impact of ranking on sales

Though the QSR chain could track sales that were coming via the food app channel, they had no way of knowing incremental organic volume driven by marketing efforts. 

One missing variable here was how much of extra sales could be attributed to improvement of QSR ranking? 

Solution: 

By combining the sales data with aggregated insights over time, we established for the QSR chain how much increase in sales they could anticipate from an increase in ranking, also knowing which changed variables led to the percentage of change increase.

So, in essence, we were able to tell the QSR chain that for each store how much sales would increase by improving ETAs, rating, ad visibility, availability, etc., enabling precise ROI calculations for each intervention they make for their stores.

Increasing sales by 3x within six months was only the beginning, and the journey of driving marketing efficiency using competitive and channel data has only just begun. 

DataWeave for QSRs

DataWeave has been working with global QSR chains, helping them drive their growth on aggregator platforms by enabling them to monitor their key metrics, diagnose improvement areas, recommend action, and measure interventions’ impact. DataWeave’s strategy eliminates the dependence on food apps for accurate data. We aggregate food app data and websites to help you with analysis and the justification of marketing spend and drive 10-15% growth.

DataWeave’s strategy eliminates the dependence on food apps for accurate data. We aggregate food app data and websites to help you with analysis and the justification of marketing spend and drive 10-15% growth.

If you want to know learn how your brand can leverage Dataweave’s data insights and improve sales, then click here to sign up for a demo

- Abhinav Gupta
Abhinav is Director of Strategic Initiatives at DataWeave and is leading the food vertical. He is working closely with several QSR chains helping them grow their business from the aggregator channels. Abhinav can be reached at abhinav.gupta@dataweave.com., 21st Sep, 2021

Brands Company E Commerce QSR Strategy



Book a Demo

  Tick this box to subscribe to our quarterly newsletter