Embodiments are directed at controlling an online advertising campaign to minimize the effective cost per event or impression. Such embodiments include receiving a desired pacing indicator and a maximum cost indicator. The desired pacing indicator indicates a desired pacing that a user would like to achieve in the campaign. The maximum cost indicator indicates a maximum cost per event or impression the user would like to stay below in the campaign. Embodiments further include monitoring bid transactions of an advertising market of the campaign. Based on monitored bid transactions, an observed pacing measurement and/or an observed cost measurement is determined. Price and allocation control signals are determined based on the desired pacing indicator, the maximum cost indicator, the observed pacing measurement, and the observed cost measurement. The control signals are utilized to adjust a bid price and/or a bid allocation for the campaign.

An example apparatus disclosed herein includes a data input interface to access different sets of initial adjustment factors, the initial adjustment factors generated to correct at least one of misattribution or non-coverage of media impressions logged by a database proprietor, a grouping controller to identify a first set of the initial adjustment factors as a set of donor adjustment factors based on (a) first audience members associated with the set of donor adjustment factors satisfying a first threshold and (b) first impressions associated with the first audience members satisfying a second threshold, and identify a second set of the initial adjustment factors as a set of recipient adjustment factors, and an imputation factor generator to generate imputation factors to correct ones of the recipient adjustment factors based on ones of the donor adjustment factors.

Methods, apparatus, systems and articles of manufacture are disclosed to generate computer-trained machine learning models to correct computer-generated errors in audience data. An example apparatus includes a query selector to select a plurality of features and a range of hyperparameters; a query generator to generate a plurality of machine learning models based on the plurality of features and the range of hyperparameters, and initiate training of the plurality of machine learning models based on demographic data in a privacy-protected cloud environment, the demographic data obtained from database proprietor user accounts corresponding to audience measurement panelists; and a model selector to select a first machine learning model from the plurality of machine learning models.

Methods and apparatus to generate audience metrics using third-party privacy-protected cloud environments. An example apparatus includes a data modifier to obtain a first matrix, the first matrix including first data indicative of entities and embeddings, the entities representative of at least one of search result clicks or videos watched, the embeddings representative of at least one of first classifications of the search result clicks or second classifications of the videos watched, generate a second matrix by reducing the first data in the first matrix to second data that satisfies a size corresponding to an input feature, and store the second matrix in first memory as the input feature, and a model generator to generate a demographic correction model based on the second matrix as the input feature, the demographic correction model to correct demographics corresponding to impressions logged in second memory.

Methods, apparatus, systems, and articles of manufacture are disclosed to adjust demographic information of user accounts to reflect primary users of the user accounts. An apparatus comprising: memory; and processor circuitry to execute instructions that causes the processor circuitry to at least: determine a first total score for a first panelist associated with a panelist user account based on at least one of a first impression score, a first age score, or a first gender score; determine a second total score for a second panelist associated with the panelist user account based on at least one of a second impression score, a second age score, or a second gender score; and in response to determining that the first total score satisfies a threshold, store demographics of the first panelist for the panelist user account.

An example apparatus disclosed herein includes a panelist detector to identify audience measurement panelists associated with database proprietor accounts, a sign-in rate calculator to determine an actual sign-in rate of the audience measurement panelists based on first impressions represented in database proprietor impressions data and second impressions represented in panel data, an adjustment factor generator to determine a first audience adjustment factor corresponding to a first sign-in rate and a second audience adjustment factor corresponding to a second sign-in rate, and a weighting controller to generate a first weighted audience adjustment factor and a second weighted audience adjustment factor by weighting the first and second audience adjustment factors by the actual sign-in rate, the adjustment factor generator to determine a signed-out adjustment factor based on the first and second weighted audience adjustment factors.

Methods and apparatus to generate audience metrics using third-party privacy-protected cloud environments. In some examples, an apparatus comprising a model generator to generate an individualization model based on truth data indicating first true users exposed to media via first panelist client devices, a model analyzer to produce user probabilities for second panelist client devices based on the individualization model, the user probabilities indicating likelihoods of second true users being exposed to media via the second panelist client devices, a data modifier to: select a user probability from the user probabilities based on an impression, the impression associated with at least one feature and a selected device from the second panelist client devices, the user probability associated with the at least one feature and the selected device, the user probability indicating a likelihood of a set of second true users being exposed to media corresponding to the impression via the selected device, and assign the impression to the set of second true users.

A method for attribution determination in a system including a plurality of vehicles having exteriors configured to convey messaging to occupants of other vehicles. Communication signals including information relating to time-stamped locations of the vehicles are received. Time-stamped location data including a plurality of the time-stamped locations is provided to a mobile location data aggregator. Anonymized identifiers associated with mobile device users within exposure zones associated with the time-stamped locations are received from the mobile location data aggregator. A control set of anonymized identifiers associated with users within a geo-fenced area but not within the exposure zones is also received. A record of conversion events is compiled. The conversion events include a first set of conversion events performed by the exposed audience and a second set of conversion events performed by the control audience. An attribution metric is determined based upon the first and second sets of conversion events.

Computer implemented sales prediction system collects data relating to events of visitors showing an interest in a client company from plural data sources, an organization module which organizes collected data into different event types and separates collected event counts in each event type between non-recent and recent events occurring within a predetermined time period, a first processing module which periodically calculates weighting for each event type based on recent and non-recent events for the event type compared to totals for other selected event types, a second processing module which periodically calculates sales prediction scores for each visitor and companies with which visitors are associated based on accumulated event data and weighting, and a reporting and data extract module which is configured to detect variation in sales prediction scores over time to identify spikes which can predict upcoming sales and to provide predicted sales information and leads to the client company.

This paper presents a practical method for measuring the impact of multiple marketing events on sales, including marketing events that are not traditionally trackable. The technique infers which of several competing media events are likely to have caused a given conversion. The method is tested using hold-out sets, and also a live media experiment for determining whether the method can accurately predict television-generated web conversions.