Multi-touch attribution implementation

hey everyone, we're really trying to level up our marketing analytics here at [my hypothetical SaaS]. right now, we're mostly stuck on last-click attribution which, as you can imagine, is just not cutting it for our increasingly complex saas funnels. we've got users interacting with us across so many different channels before they convert, and last-click completely ignores all those crucial touchpoints that influenced their decision. it's giving us a really skewed view of what's actually driving our growth and where we should be allocating our budget. the real challenge we're grappling with is implementing true multi-touch attribution. we want to move beyond simple rule-based models and dive into something more algorithmic or data-driven, but the complexity is pretty daunting. integrating disparate data sources – think ad platforms, crm, website analytics, product usage data – and then figuring out how to accurately assign fractional credit across non-linear conversion paths is proving to be a massive headache. we're looking at things like shapley values or markov chains, but the practicalities of setting these up, preparing the data, and then validating the models in a real-world saas context feels like a huge undertaking. does anyone have practical advice on frameworks or tools that have worked well for them in building out a robust multi-touch attribution model? especialy keen on insights regarding data preparation and model selection. anyone faced this before?