Introduction
Ever wondered if the images captured from the space could also be utilized for improving your business brand image? The outdoor advertising planners use the remotely sensed images in their workflow to choose where to place the billboard advertisements. We can better comprehend, interpret, and analyze the issue of where to put billboards in a city by combining remotely sensed data with the derived data products in a Geographic Information System (GIS). This blog post deals with the billboard placement challenge within any city.
What is billboard advertising?
Billboard advertising is one of the most prominent forms of marketing. Every now and then, while driving through a city, we come across out-of-home advertising boards. This is particularly pertinent to smart cities since they are all about connections—that is, people, things, data, and processes. The role of marketing is crucial in smart cities.
Do billboards still serve as effective forms of advertising?
The purpose of outdoor billboard advertisements may seem irrelevant in the digital age, where ads are strategically placed based on our online content consumption habits. However, studies have found that the overall reach of billboard advertisements is quite impressive. As per the Outdoor Advertising Association of America (OAAA), 71% of travelers look at the roadside billboards. Additionally, billboards have high recall and retention rates, indicating how memorable they are to the audience. Billboard advertisements have been successful in not only driving the attention of the onlookers but also in helping to attain benefits further along the advertising model chain of Attention-Interest-Desire-Action-Retention (AIDAR). It has the capability to reach multiple target groups. Apart from being cost-effective, billboard advertisements complement the digital outreach campaign of any marketing strategy.
Why is the optimum placement of billboards important?
In conventional marketing, there are four Ps: product, price, place, and promotion. Place is an important component in marketing; hence, geographic location plays an important role.The billboard advertisement has a brief window of time to capture attention; therefore, utilizing the audience’s photographic memory is essential. For advertising effectiveness, the location of the billboard is found to be more relevant than the visual appearance of the billboard. An inappropriately placed billboard is a waste of investment.
What is the general trend in the placement of billboards?
Generally, billboards are placed at locations where the sight is easy and in quite a lot of numbers, such as city centers and busy highways. As a general rule, it is possible to believe that the city center is the ideal place for billboard advertisements. Building mathematical models, conducting travel surveys, or manually calculating traffic volume could all be involved in such tasks.
Where’s the challenge?
Since not every advertisement can be positioned in a city center, planners of advertisements will need to choose the best sites outside of these areas. The biggest challenge here is how to accurately quantify the effectiveness of a billboard after placement. For this, traffic volume on the nearby road segment and possible trajectory pattern analysis need to be carried out. A few researchers have also tried visibility analysis from the perspective of passengers for product placement effectiveness. In the event that the efficacy of a billboard location can be measured, the next problem is determining which of all the potential locations for the billboard is the best. In addition to choosing appropriate locations, users also want to see the similarities and differences between the candidate locations they have chosen.
The geospatial approach for the billboards optimum location problem
Placing the billboards appropriately is the spatial optimization site selection problem. A brief summary of all the datasets that can be used to solve such problems and the methods used to do so is provided in the sections that follow.
Inputs
Remotely sensed images: There is a large selection of holistic view images that can be used as a foundation for data generation, analysis, and visualization.
Products based on digital photogrammetry: Digital photogrammetry is used to create digital elevation models and 3D city models for external morphology.
Road network data: While there are numerous ways to obtain this information, satellite imagery is the primary source used to prepare or update the road dataset. Using the taxi trajectories data can also lead to a road network layer. Using a technique called map matching, the trajectory data is transformed into road network data. Road sequence data is created here by converting GPS sequence data. Another derived product that can be produced is road density.
Traffic volume data: This dataset is required for knowing the average speed of traffic. Their distribution across day slots also matters, as do their origins and destinations. Traffic density is also calculated. In addition, traffic prediction problems constitute an entire field. There are three types of predictions: traffic status, flow, and demand predictions.
Features of the surrounding area: It includes information about the number of buildings and notable points of interest (POIs), protected and forbidden areas, monuments, museums, historical sites, and tourist destinations, educational institutions, medical facilities, pharmacies, and places to stay, among other things.
Demographic data: It encompasses the characteristics of the audience for targeted advertising. Their shifting also matters.
Population density or population distribution data: LandScan gridded population data at a 1×1 km grid is mostly suitable for this purpose.
Billboard point data: It comprises the spatial distribution of candidate billboard locations. Here, the budget for each billboard can also be considered.
Analysis
The two well-known spatial optimization models are as follows:
Location Set Covering Problem (LSCP): This type of analysis aims at finding the minimum number of billboard locations that can ensure complete coverage. However, for the billboard location problem, this method is often not used.
Maximum Covering Location Problem (MCLP): Here it is tried to cover a maximum possible area with the given fixed number of billboards to be placed. This is the preferred model when it comes to optimizing location for the placement of billboards. In this method, first a grid of spatial units is prepared (e.g., gridded point data spaced at 1 km from each other). Then, traffic flow and other characteristics are calculated. Then, this data is brought to the road segments. Here, roads are marked by the type of road (primary, secondary, or tertiary). Now the billboard grid points are taken, and spatial join is performed with the traffic-attributed road segment data falling within the 1 km range, after which the model is run. This model searches for the billboard location that covers the maximum traffic volume. This is a case of COPs (combinatorial optimizing problems), where now neural network-based deep reinforcement learning is used to address challenges like the complexity of real world situations and the limited availability of data. This approach has shown promising results in finding the optimal solution for billboard placement.
Obstacles to using the geospatial data-driven approach
The cost of acquiring data is one of the main obstacles. Furthermore, when datasets are sufficiently large, running times are high. Alternatively, high computational power is required. Expert help is still required when things become complex.
Geospatial future of billboard advertising
The billboard placement problem falls under geographically enabled business intelligence applications. Modern web map based data visualization techniques, which some large outdoor advertising companies are already using, are among the futuristic solutions in this field. Here, a portal of web maps is prepared, where the clients can themselves select the appropriate locations. The portal has provisions for various datasets, like traffic volume and demographic data, and the model is embedded in a few clicks. The map can display the suitable candidate locations as per the user-supplied information. A geospatial big data computing framework is anticipated to be utilized in the future for smart cities.
References
Fortenberry, J. L., & McGoldrick, P. J. (2020). Do Billboard Advertisements Drive Customer Retention?: Expanding the “AIDA” Model to “AIDAR.” Journal of Advertising Research, 60(2), 135–147. https://doi.org/10.2501/JAR-2019-003
Khargharia, H. S., Rehman, M. H. ur, Banerjee, A., Montori, F., Forkan, A. R. M., & Jayaraman, P. P. (2023). Towards Marketing 4.0: Vision and Survey on the Role of IoT and Data Science. Societies, 13(4), Article 4. https://doi.org/10.3390/soc13040100
Liu, D., Weng, D., Li, Y., Bao, J., Zheng, Y., Qu, H., & Wu, Y. (2017). SmartAdP: Visual Analytics of Large-scale Taxi Trajectories for Selecting Billboard Locations. IEEE Transactions on Visualization and Computer Graphics, 23(1), 1–10. https://doi.org/10.1109/TVCG.2016.2598432
Yuan, H., & Li, G. (2021). A Survey of Traffic Prediction: From Spatio-Temporal Data to Intelligent Transportation. Data Science and Engineering, 6(1), 63–85. https://doi.org/10.1007/s41019-020-00151-z
Zhong, Y., Wang, S., Liang, H., Wang, Z., Zhang, X., Chen, X., & Su, C. (2024). ReCovNet: Reinforcement learning with covering information for solving maximal coverage billboards location problem. International Journal of Applied Earth Observation and Geoinformation, 128, 103710. https://doi.org/10.1016/j.jag.2024.103710
