Non-Exclusive Notifications for Ride-Hailing at Lyft II: Simulations and Marketplace Analysis

Abstract

Ride-hailing platforms increasingly face uncertain driver acceptance, which makes traditional one-to-one 'exclusive dispatch (ED)' less efficient: rejections and timeouts force sequential retries and lengthen rider wait times, which in turn creates friction in the marketplace. 'Non-exclusive dispatch (NED)' mitigates this friction by broadcasting a request to multiple drivers in parallel. While NED can reduce latency, it introduces new design challenges -- most notably, how to choose notification sets and how to resolve driver contention (when multiple drivers accept the same ride). In this paper -- the second in a two-part collaboration with Lyft -- we develop a theoretically grounded framework to evaluate the long-run performance and marketplace effects of transitioning from ED to NED. We bridge theory and practice by combining (i) an optimization model that formulates NED as a constrained welfare maximization problem with (ii) large-scale discrete-event simulations on proprietary Lyft traces and (iii) a stylized macroscopic equilibrium model. Across simulation and equilibrium analysis, we find that NED improves key fulfillment metrics relative to ED: it reduces match time (and hence rider reneging) while increasing both the number and the average quality of completed matches. We also quantify the speed--quality trade-off between two common contention resolution rules, 'First-Accept' and 'Best-Accept': First-Accept maximizes speed and throughput, whereas Best-Accept is required to maximize per-match quality. Finally, we show that slightly conservative notification heuristics can improve long-run efficiency by avoiding excessive locking of high-value drivers and preserving future availability.

Figures (16)

• Figure 1: Flowchart illustrating the simulation lifecycle from a rider's perspective. (Note: Under BA if the highest-ranked driver accepts prior to the 7th cycle, the match is finalized immediately.)
• Figure 2: ED vs NED with optimal packing ($\boldsymbol{U=3, \theta=0}$) under FA/BA; The panels report the histogram of average match score, average rider match time, and total number of finalized matches across sampled instances.
• Figure 3: Sensitivity analysis of the proposed algorithms. The red markers ($\boldsymbol{U=1}$) correspond to the exclusive-dispatch baseline and serve as a common starting point for FA (blue) and BA (green) strategies. Notably, the ED+ and the rejection-aware heuristic exhibit minimal variance with respect to $\boldsymbol{\theta}$ and $\boldsymbol{U}$, because ED+ is limited to notifying at most two drivers per request and the heuristic rarely selects more than two drivers.
• Figure 4: Performance landscape of our algorithms in Appendix \ref{['sec:Algs']} across different configurations. Left: average score vs. average match time. Right: total match count vs. average match time. Each point corresponds to a unique combination of packing algorithm (marker), contention rule (FA vs. BA), $\boldsymbol{U\in\{2,3,4,5\}}$, and $\boldsymbol{\theta\in\{0,\dots,0.5\}}$. The exclusive-dispatch baseline ($\boldsymbol{U=1}$) is highlighted in red.
• Figure 5: Performance metrics (average score, average match time, match count) for the optimal algorithm, i.e., NED OPT, under $k$-accept strategies with $\boldsymbol{U=5,\theta=0}$, for $\boldsymbol{k\in\{1,2,\ldots,5\}}$.

Theorems & Definitions (5)

• Proposition 4.1: FA Equilibrium
• Example 4.1
• Proposition 4.2: BA Equilibrium
• Theorem 4.3
• Proposition 4.4