We have a number of algorithms that can derive the optimal spending strategy to target voter turnout, swing voters, and voting precincts.
Maximize campaign budgets by utilizing our campaign spending strategy options.
We translate campaign parameters from words, time, and dollars to mathematical terms. Then set up the optimization with these parameters "uncertaintized" and constraints applied.
We utilize a number of well-proven algorithms to derive or validate optimal campaign spending strategies. Both gradient-based and more heuristic methods can be used.
Once an optimal campaign spending strategy is developed we like to re-run the process a number of times to show how sensitive or robust the strategy is to uncertain inputs. This produces a more likely range of outcomes rather than a single point output.
Our calculated levels of confidence indicate whether the derived optimal spending strategy is a "thread the needle" strategy or a more forgiving "close enough" strategy. This informs campaign decisions going forward.
We give you options on how to spend campaign resources to maximize your chance of winning the election.
A political campaign is always a lot of trade-offs and difficult decisions to make. We have mathematical algorithms that can find the optimal ways to spend cmpaign finances and gain votes. Our algorithms pinpoint which precincts to target and which precincts to spend less on.
Our algorithms allow you to specify hard constraints on the campaign strategy optimization process. A hhard constraint on spending? Done! A hard constraint on precincts to ignore? Done!
Our optimal campaign strategy algorithms run quickly for moderate-sized efforts such as state sensator or representative. Thus we can quickly run a number of scenarios for new data, changing emphasis, and/or changing campaign budgets.
There is no certainty in a political campaign. Thus our algorithms deal with the uncertainty of a political campaign concerning input data, campaign decisions, or changing times. And we can offer confidence levels in our results as a function of the uncertainty.