Background: Debates on the future of work foreground digital skills, automation, and AI, yet empirical evidence on digital skill demand is still disproportionately derived from advanced-economy job advertisements and occupation-centric inference. This study reframes digitalization in SME-dominated African economies as a work-design problem, observable in how firms themselves specify “digital work to be done” as tasks or projects rather than as formal roles. Methods: We analyze an anonymized dataset of N = 250 African SMEs that provided standardized disclosures of digital domains, task deliverables, preferred work modality, intended contributor count, and supervision/co-investment signals. Using a descriptive-analytical strategy, we code digital domains, classify task orientation (task-specific vs project-based), model demand intensity via intended contributors, examine modality as a signal of codifiability, and extract latent demand archetypes via topic modeling. Results: Digital work demand is predominantly articulated as tightly scoped, task-specific requests, with project-based articulation emerging more selectively where integration and learning requirements rise. Demand intensity is highly skewed: the median SME seeks a very small team (often 1–3 contributors), while a minority articulate large-scale initiatives (tail up to ~50). Work modality is not merely logistical: remote and hybrid preferences dominate overall, while more complex, data-intensive and project-based work shifts toward hybrid/in-person, consistent with higher coordination and tacit knowledge needs. Topic modeling further reveals five recurring demand archetypes centered on applied, operational digitalization rather than frontier experimentation. Conclusions: A task-based, demand-side lens surfaces how capability constraints and absorptive capacity shape not only whether SMEs digitalize, but how they architect digital work, with implications for future-of-work measurement, SME support programs, and inclusive digital transformation theory.