Introduction
In the run-up to the forthcoming publication of my new book, 'Patterns in Brand Monitoring: A Scientific Approach to Brand Protection Analysis', I conduct a retrospective overview of a series of brand protection analysis case studies involving visualisations of a range of types of data.
Data analysis and visualisation can have a number of applications in brand protection, including: the identification of trends and patterns of activity and information relating to new or emerging areas of concern; providing insights into serial or high-activity bad actors or the associated focuses of infringing activity, thereby allowing prioritised monitoring or enforcement initiatives with greater efficiency; clustering together related findings and collating evidence of bad-faith activity; determining the financial impacts of intellectual property crime and allowing return-on-investment analyses; and benchmarking brands against their peers.
Data visualisation case studies
1. Profiling a scam package-tracking website[1,2]
This case study involves the profiling of a single central scam website associated with a large, coordinated brand impersonation attack targeting large numbers of well-known brands, using many thousands of lookalike sites. The scam website solicited for payments, purportedly to authorise the delivery of items ordered through the lookalike sites, utilising a range of personalised, recipient-specific pages, with URLs differing from each other only by the numeric string (the 'ID-number') at the end of each link.
Figure 1.1: Heat map showing the total numbers of active scam pages in each 'sub-block' of 50 adjacent ID-numbers (with darker shades indicating greater utilisation of the available set of ID-numbers), for the first block of 100,000 possible ID-numbers
Figure 1.2: Daily mean payment requested per individual scam page, during the full duration of the period of use of the scam site
Figure 1.3: 'Timeline' view showing the number of pages on which each of the top 80 overall most frequently used e-mail addresses (obfuscated) was utilised, within each calendar month
2. The GameStop saga[3]
This case study provides an illustration of how high-profile real-world events can trigger associated spikes in related infringements, by bad actors looking to take advantage of increased levels of interest and search volumes. The event in question was the Reddit campaign to boost the share price of US retailer GameStop in January 2021, with the study focusing on the daily numbers of brand-related domain registration in the period surrounding the campaign. The resulting analysis showed how the numbers of registrations tracked the company's share price extremely closely.
Figure 2.1: Daily numbers of registrations of domain names containing 'gamestop', compared with the daily high share price for the organisation
3. Domain-name availability[4,5]
This case study looks to quantify the observation that the availability of short, memorable, unregistered domain names across popular extensions (TLDs) is increasingly running low, a trend which is pushing brand owners towards the use of novel or invented brand names, alternative TLDs, dot-brands, or other technologies such as blockchain domains. The study uses zone-file analysis to determine the proportion of all possible domain names of each length (up to 6 characters) which are already registered, across the top 40 extensions.
Figure 3.1: Proportion of the set of all possible domain names which are already registered, for each of the top 40 gTLDs (by total number of registered one- to six-character domains), as a function of SLD length (n characters)
4. Subdomain discovery[6]
This case study involves the test of a series of methods intended to identify as many existing subdomains as possible across arbitrary third-party domains (websites), using the top 50 most popular sites as an example. Discovery of subdomains on arbitrary websites is a key component of brand monitoring programmes because of the potential for brand abuse, but is traditionally a difficult problem to solve because of the lack of publicly-accessible data sources akin to domain name zone files. It is therefore instructive to test and develop discovery algorithms and seek insights into keyword, length and naming patterns in subdomain usage.
Figure 4.1: Distribution of subdomain lengths (in characters) and numbers of levels across the dataset, by number of instances
5. The relationship between brand value and brand prominence[7,8]
This case study involves the calculation of online prominence for each of the top 100 most valuable global brands, using a proprietary algorithm considering the number and prominence of the mentions of each of the brands, based on analysis of a generically-sampled set of webpages. A comparison of the prominence scores for each of the brands with their brand values shows no straightforward relation between the two, though some trends (such as the disproportionate high prominence of social media, search and technology brands, and the low prominence of luxury brands) are evident.
Figure 5.1: Comparison of overall prominence score with brand value for the top 100 brands, split by industry area (showing low ends of prominence / value axes only)
6. Real-world distribution of infringing goods[9,10]
This case study looks to tie together online and offline activity relating to the trade in infringing goods. It considers the locations of interception in the UK of incoming goods from three countries identified as the top points of origin, based on a case study of partnerships with customs and law enforcement for three key brands in different industry areas (clothing, food, and physical goods manufacturing).
Figure 6.1: Heat map showing the frequency with which locations in the UK have been associated with the physical interception of infringing goods originating from China (for three brands in the clothing, food, and physical goods manufacturing industries)
7. Colour similarity measurement[11]
This case study looks at a set of protected colour marks (or colours featured as components of more complex marks) and considers a framework for quantifiably measuring the difference between each pair of colours, based on the geometric distance between them in RGB (red-green-blue) space[12]. Such metrics could form the basis of a more robust approach to quantifying the difference between marks, as may be relevant to the decision-making process for disputes[13].
Figure 7.1: Matrix showing the separation (in RGB units) between each pair of colours in the set of marks
References
[1] 'Patterns in Brand Monitoring' (D.N. Barnett, Business Expert Press, 2025), Case Study 8.6: 'Profiling a scam package-tracking website'
[2] https://www.iamstobbs.com/tracking-the-tracker-ebook
[3] 'Patterns in Brand Monitoring' (D.N. Barnett, Business Expert Press, 2025), Case Study 8.1: 'The GameStop Saga'
[4] 'Patterns in Brand Monitoring' (D.N. Barnett, Business Expert Press, 2025), Section 9.3.2: 'Availability of short domains across the gTLD landscape'
[5] https://www.iamstobbs.com/availability-of-domains-ebook
[6] https://circleid.com/posts/20240528-exploring-the-domain-of-subdomain-discovery
[7] 'Patterns in Brand Monitoring' (D.N. Barnett, Business Expert Press, 2025), Case Study 10.1: 'Online prominence and sentiment of the top 100 most valuable global brands'
[8] https://www.iamstobbs.com/online-brand-prominence-and-sentiment-ebook
[9] 'Patterns in Brand Monitoring' (D.N. Barnett, Business Expert Press, 2025), Chapter 15: 'Links to offline data'
[10] https://www.iamstobbs.com/opinion/tracking-the-uk-trade-in-fakes-counterfeit-hotspots
[12] https://circleid.com/pdf/similarity_measurement_of_marks_part_1.pdf
This article was first published on 24 December 2024 (amended 25 December 2024) at:
https://www.linkedin.com/pulse/brand-protection-data-beautiful-david-barnett-c66be/
No comments:
Post a Comment