1 Reply Latest reply on Dec 15, 2016 2:00 PM by Neil Wright

    SFDC-MKTO Standard Sync Volumes & Behavior

    Kelly Shoemake

      I'm trying to find documentation to understand the standard volume of data that can be synced between Salesforce and Marketo. I've run into a flurry of issues with the sync recently and it is difficult to determine if this is what I should expect or if there's something critically wrong.

       

      The most I've been able to find are items published by community members or on a tangent topic, such as

      How often are SFDC campaigns updated in Marketo? Re: How often are SFDC campaigns updated in Marketo?

      "Marketo SFDC Sync can process up to 20k records per hour."

      Changes to Marketo Salesforce Sync – Under The Hood

      "The backfill process will push data at roughly the same rate as the “Push Updates” rate we get now in the regular sync—somewhere between 7k to 12k records per hour."

       

      I have a case I opened recently where I was told it by Marketo support that it would take 7 hours to process 14,000 records, which is far off from the two items I listed above.

       

      I'm also trying to understand if the standard behavior for updates is to push all updates at once or break it out. During this same incident, I observed that split updates were occurring. To explain, I had a batch campaign with two change data values in a flow. If these were scheduled to update on 20,000 records, the first change data value pushed for all records, then the second pushed for the same 20,000 records again, making the sync work as if it updated 40,000 records.

       

      I've heard that the sync flinches when there are many updates to a record at once that can cause the update speed per record decrease, but again I haven't found any documentation to understand how many updates are considered reasonable versus pushing system limitations.

        • Re: SFDC-MKTO Standard Sync Volumes & Behavior
          Neil Wright

          I've found it extremely difficult to judge Marketo's ability (or lack of) in processing large numbers of records, whether it's processing data changes, adding to campaigns, sync-ing to SFDC, or running through triggered or waterfall campaigns. Along with a lack of campaign prioritization or controllable flows, the slow bulk processing capabilities of Marketo is one my least favorite constraints.

           

          I know it's not the answer you want to hear, but it very much depends on your systems (Marketo and SFDC) as to the throughputs you can achieve. To give you an idea, though, I have done a lot of bulk processing because we have migrated from a non-Marketo system to Marketo then from our original Marketo (with a custom integration to a CRM) to a "classic" Marketo-SFDC set up - a lot of bulk processing! What I have found is the following:

           

          Data importing (avoiding a system backlog): about 2,000 per hour

          First time syncing of large numbers of leads to a SFDC campaign: we get about 500 per hour

          Data change (i.e. changing the value in a single Marketo field): about 6,000-12,000 per hour

           

          In your scenario (2 data values being changed, then sync'd), it would appear to me that making both changes first then syncing both updates at once would make sense, but it may be that someone with better knowledge of how/why the sync works may argue.

           

          Hope this helps.