Browse Prior Art Database

Extraction of Meaningful Marketing Campaign KPIs from Online Chatter Disclosure Number: IPCOM000238845D
Publication Date: 2014-Sep-22
Document File: 2 page(s) / 31K

Publishing Venue

The Prior Art Database


Measuring the effectiveness of marketing campaigns across various earned, owned and paid media channels is one of the most challenging tasks for today?s brand marketers. With a limited campaign budget and numerous channel possibilities for campaign execution, the assessment of marketing effectiveness per channel becomes an essential requirement for deriving ?optimal" marketing strategies and more flexible cross-channel budget allocations. Measurement of marketing campaigns? effectiveness usually relies on a combination of key performance indicators (KPIs), used for assessing various aspects of marketing outcomes (e.g., response leads generation, brand image, etc). KPI measurements can be then aligned against various marketing objectives set by the brand?s executive level, e.g., return on marketing investment (ROMI), cost per lead, lift in sales or brand?s image, etc. Two main pitfalls can be identified in existing solutions. First, the relatively simple KPIs (e.g., number of impressions, sentiments, sharings, etc) that are currently extracted by existing solutions fail to provide satisfactory evidence for a campaign?s performance. The main cause of such failure is due campaign independent factors that are currently ignored by most existing solutions, such as the brand?s general popularity, which may govern extracted KPIs ?readings" and add unobserved ?noise". Second, existing KPIs do not provide sufficient correspondence with marketing business goals such as sales lift, ROMI, etc. For KPIs to be effective, they must provide clear evidence on the campaign?s ability to generate new leads or improve the brand?s image during its execution Extraction of novel marketing performance KPIs from online chatter that provide statistical significant indication of the success of campaigns. Two main dimensions are captured by the KPIs: 1. The ability to focus the brand's potential audience on the campaign. 2. The ability of the campaign in generating (future) leads.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 01 of 2

Extraction of Meaningful Marketing Campaign KPIs from Online Chatter

The availability of online chatter data collected from a stream of user generated content (UGC) is now assumed.

For a given any UGC update, we now denote B,C,L the event labels that correspond to mentions of the brand, campaign and a lead in the update respectively.

Similarly, we denote nB,nC,nL the absence of such event labels in UGC updates. Each UGC update's content is analyzed and and pass three different classifiers (e.g., MNB) that tag the update as follows:
1. B/nB: The UGC update mentions the brand (e.g., the update contains the name of the brand or one of its product names)

2. C/nC: The UGC update mentions the campaign (e.g., the update contains the slogan of the campaign or mentions the name of the campaign presenter).

3. L/nL: The UGC update contains a lead indicator (e.g., the update includes wishful thinking about future purchase of the brand's products)

    Given a (random) sample of UGC updates that were collected in some time period, we now propose four novel marketing campaign KPIs extracted from online chatter that provide statistical
significant indication on a campaign's success:

Campaign Focus: This KPI captures the relative focus a given campaign receives from the brand's targeted audience.

Borrowing ideas from the association rules mining literature, such KPI is measured as the confidence of the association rule: campaign ==> brand, defined as the relative number of brand's impressions that

also mention the campaign. The higher the confidence is, the more we attribute the campaign for influencing the brand's targeted audience beyond its general popularity.

Formally, this KPI is measured as: P(C,B), and derived using maximum likelihood estimation given sampled UGC updates.

Leads Attribution: This KPI aims at providing hints on the most essential question that bothers today's marketers: "can the campaign